• English
• Deutsch

• University of Freiburg
• Faculty of Engineering
• Department of Computer Science

Publications by author

[Johannes Aldinger]    [Dr. Yusra Alkhazraji]    [Christian Becker-Asano]    [Gregor Behnke]    [Rolf-David Bergdoll]    [Michael Brenner]    [Markus Büttner]    [Patrick Caspari]    [Yannis Dimopoulos]    [Christian Dornhege]    [Christoph Dornheim]    [Thorsten Engesser]    [Patrick Eyerich]    [Florian Geißer]    [Jens-Steffen Gutmann]    [Moritz Göbelbecker]    [Wolfgang Hatzack]    [Malte Helmert]    [Andreas Hertle]    [Jörg Hoffmann]    [Julien Hué]    [Thomas Keller]    [Alexander Kleiner]    [Jana Koehler]    [Sebastian Kupferschmid]    [Christian Köhler]    [Felix Lindner]    [Johannes Löhr]    [Robert Mattmüller]    [Bernhard Nebel]    [Manuela Ortlieb]    [Marco Ragni]    [Jochen Renz]    [Jussi Rintanen]    [Gabriele Röger]    [Alexander Schimpf]    [Tim Schulte]    [Alexander Scivos]    [Jan-Georg Smaus]    [David Speck]    [Dali Sun]    [Sebastian Trüg]    [Thilo Weigel]    [Bruno Welsch]    [Matthias Westphal]    [Jens Witkowski]    [Benedict Wright]    [Stefan Wölfl]    [Dapeng Zhang]   

(Show all abstracts) (Hide all abstracts)

Johannes Aldinger

• Daniel Kuhner, Lukas D.J. Fiederer, Johannes Aldinger, Felix Burget, Martin Völker, Robin T. Schirrmeister, Chau Do, Joschka Boedecker, Bernhard Nebel, Tonio Ball and Wolfram Burgard.
  A service assistant combining autonomous robotics, flexible goal formulation, and deep-learning-based brain–computer interfacing.
  Robotics and Autonomous Systems 116, pp. 98-113. 2019.
  (Show abstract) (Hide abstract) (Online; DOI)
  

  As autonomous service robots become more affordable and thus available for the general public, there is a growing need for user-friendly interfaces to control these systems. Control interfaces typically get more complicated with increasing complexity of robotic tasks and environments. Traditional control modalities such as touch, speech or gesture are not necessarily suited for all users. While some users can make the effort to familiarize themselves with a robotic system, users with motor disabilities may not be capable of controlling such systems even though they need robotic assistance most. In this paper, we present a novel framework that allows these users to interact with a robotic service assistant in a closed-loop fashion, using only thoughts. The system is composed of several interacting components: a brain–computer interface (BCI) that uses non-invasive neuronal signal recording and co-adaptive deep learning, high-level task planning based on referring expressions, navigation and manipulation planning as well as environmental perception. We extensively evaluate the BCI in various tasks, determine the performance of the goal formulation user interface and investigate its intuitiveness in a user study. Furthermore, we demonstrate the applicability and robustness of the system in real-world scenarios, considering fetch-and-carry tasks, close human–robot interactions and in presence of unexpected changes. As our results show, the system is capable of adapting to frequent changes in the environment and reliably accomplishes given tasks within a reasonable amount of time. Combined with high-level task planning based on referring expressions and an autonomous robotic system, interesting new perspectives open up for non-invasive BCI-based human–robot interactions.

• Daniel Kuhner, Johannes Aldinger, Felix Burget, Moritz Göbelbecker, Wolfram Burgard and Bernhard Nebel.
  Closed-Loop Robot Task Planning Based on Referring Expressions.
  In Proceedings of the 2018 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS 2018), pp. 876-881. 2018.
  (Show abstract) (Hide abstract) (PDF)
  

  Increasing the accessibility of autonomous robots also for inexperienced users requires user-friendly and high-level control opportunities of robotic systems. While automated planning is able to decompose a complex task into a sequence of steps which reaches an intended goal, it is difficult to formulate such a goal without knowing the internals of the planning system and the exact capabilities of the robot. This becomes even more important in dynamic environments in which manipulable objects are subject to change. In this paper, we present an adaptive control interface which allows users to specify goals based on an internal world model by incrementally building referring expressions to the objects in the world. We consider fetch-and-carry tasks and automatically deduce potential high-level goals from the world model to make them available to the user. Based on its perceptions our system can react to changes in the environment by adapting the goal formulation within the domain-independent planning system.

• Johannes Aldinger and Bernhard Nebel.
  Interval Based Relaxation Heuristics for Numeric Planning with Action Costs.
  In KI 2017:Advances in Artificial Intelligence (KI 2017), pp. 15-28. Springer International Publishing 2017.
  (Show abstract) (Hide abstract) (PDF)
  

  Many real-world problems can be expressed in terms of states and actions that modify the world to reach a certain goal. Such problems can be solved by automated planning. Numeric planning supports numeric quantities such as resources or physical properties in addition to the propositional variables from classical planning. We approach numeric planning with heuristic search and introduce adaptations of the relaxation heuristics h_max, h_add and h_FF to interval based relaxation frameworks. In contrast to previous approaches, the heuristics presented in this paper are not limited to fragments of numeric planning with instantaneous actions (such as linear or acyclic numeric planning tasks) and support action costs.

• F. Burget, L.D.J. Fiederer, D.Kuhner, M.Völker, Johannes Aldinger, R.T. Schirrmeister, C.Do, J.Boedecker, Bernhard Nebel, T.Ball and W.Burgard.
  Acting Thoughts: Towards a Mobile Robotic Service Assistant for Users with Limited Communication Skills.
  In Proceedings of the European Conference on Mobile Robotics (ECMR 2017), pp. 385-390. 2017.
  (Show abstract) (Hide abstract) (PDF)
  

  As autonomous service robots become more affordable and thus available also for the general public, there is a growing need for user friendly interfaces to control the robotic system. Currently available control modalities typically expect users to be able to express their desire through either touch, speech or gesture commands. While this requirement is fulfilled for the majority of users, paralyzed users may not be able to use such systems. In this paper, we present a novel framework, that allows these users to interact with a robotic service assistant in a closed-loop fashion, using only thoughts. The brain-computer interface (BCI) system is composed of several interacting components, i.e., non-invasive neuronal signal recording and decoding, high-level task planning, motion and manipulation planning as well as environment perception. In various experiments, we demonstrate its applicability and robustness in real world scenarios, considering fetch-and-carry tasks and tasks involving human-robot interaction. As our results demonstrate, our system is capable of adapting to frequent changes in the environment and reliably completing given tasks within a reasonable amount of time. Combined with high-level planning and autonomous robotic systems, interesting new perspectives open up for non-invasive BCI-based human-robot interactions.

• Johannes Aldinger and Bernhard Nebel.
  Extended Abstract: Interval Based Relaxation Heuristics for Numeric Planning with Action Costs.
  In Proceedings of the 10th International Symposium on Combinatorial Search (SoCS 2017), pp. 155-156. 2017.
  (Show abstract) (Hide abstract) (PDF)
  

  We adapt the relaxation heuristics h_max, h_add and h_FF to interval based numeric relaxation frameworks, combining them with two different relaxation techniques and with two different search techniques. In contrast to previous approaches, the heuristics presented here are not limited to a subset of numeric planning and support action costs.

• Johannes Aldinger, Robert Mattmüller and Moritz Göbelbecker.
  Complexity Issues of Interval Relaxed Numeric Planning.
  In KI 2015: Advances in Artificial Intelligence (KI 2015). 2015.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  Automated planning is computationally hard even in its most basic form as STRIPS planning. We are interested in numeric planning with instantaneous actions, a problem that is not decidable in general. Relaxation is an approach to simplifying complex problems in order to obtain guidance in the original problem. We present a relaxation approach with intervals for numeric planning and show that plan existence can be decided in polynomial time for tasks where dependencies between numeric effects are acyclic.

• Johannes Aldinger, Robert Mattmüller and Moritz Göbelbecker.
  Complexity Issues of Interval Relaxed Numeric Planning.
  In Proceedings of the ICAPS-2015 Workshop on Heuristic and Search for Domain-Independent Planning (HSDIP 2015). 2015.
  Superseded by the KI 2015 paper of the same name..
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  Automated planning is a hard problem even in its most basic form as STRIPS planning. We are interested in numeric planning tasks with instantaneous actions, a problem which is not even decidable in general. Relaxation is an approach to simplifying complex problems in order to obtain guidance in the original problem. In this paper we present a relaxation approach with intervals for numeric planning and discuss the arising complexity issues.

• Johannes Löhr, Johannes Aldinger, Stefan Winkler and Georg Willich.
  Automated Planning for Earth Observation Spacecraft under Attitude Dynamical Constraints.
  In Jahrbuch der Deutschen Gesellschaft für Luft- und Raumfahrt (DGLR2013). 2013.
  (Show abstract) (Hide abstract) (PDF)
  

  Agile Earth observation missions continuously require a large amount of planning during the spacecraft's observations. Beside priorities of the observation sites, especially the agility constraints of the satellite are important to be taken into account during the planning process. This is due to the body-fixed instrument's line of sight, requiring the whole satellite to point to the observation sites while scanning. Scanning a sequence of observation sites leads to complex slew maneuvers which must not exceed the satellite's actuator capacities, attitude constraints or maximum angular rates. Additionally, the regions of interest may change over time, making it necessary to adapt and optimize the observation sequence continuously. An automated process is required to efficiently handle this task. We present a planning algorithm to sequence an arbitrarily distributed set of observation patches to a feasible observation plan, considering priority criteria of the observation sites and agility constraints of the satellite.

• Johannes Aldinger and Johannes Löhr.
  Planning for Agile Earth Observation Satellites.
  In Proceedings of the ICAPS-2013 Workshop on Planning in Continuous Domains (PCD), pp. 9-17. 2013.
  (Show abstract) (Hide abstract) (PDF)
  

  Agile Earth observation satellites are satellites orbiting Earth with the purpose to gather information of the Earth's surface by slewing the satellite toward regions of interest. Constraints arise not only from dynamical and kinematic aspects of the satellite and its sensors. Regions of interest change over time and bad weather can conceal important observation targets. This results in a constant need to replan the satellite's tasks and raises the desire to automatize this planning process. We consider the Earth observation problem with the help of the module extension of the numerical planning system Temporal Fast Downward. Complex satellite slew maneuvers are calculated within modules, while the planner selects and schedules the regions to be scanned. First results encourage deeper research in this area so that forthcoming satellite space missions can draw on automated planning to improve the performance of agile Earth observation tasks.

Dr. Yusra Alkhazraji

• Dr. Yusra Alkhazraji and Martin Wehrle.
  Sleep Sets Meet Duplicate Elimination.
  In Proceedings of the Ninth Annual Symposium on Combinatorial Search (SoCS 2016) (SoCS 2016). 2016.
  (Show abstract) (Hide abstract) (PDF)
  

  The sleep sets technique is a path-dependent pruning method for state space search. In the past, the combination of sleep sets with graph search algorithms that perform duplicate elimination has often shown to be error-prone. In this paper, we provide the theoretical basis for the integration of sleep sets with common search algorithms in AI that perform duplicate elimination. Specifically, we investigate approaches to safely integrate sleep sets with optimal (best-first) search algorithms. Based on this theory, we provide an initial step towards integrating sleep sets within A* and additional state pruning techniques like strong stubborn sets. Our experiments show slight, yet consistent improvements on the number of generated search nodes across a large number of standard domains from the international planning competitions.

• Jendrik Seipp, Florian Pommerening, Silvan Sievers, Martin Wehrle, Chris Fawcett and Dr. Yusra Alkhazraji.
  Fast Downward Aidos (planner abstract).
  In the 1st Unsolvability International Planning Competition (IPC 2016). 2016.
  (Show abstract) (Hide abstract) (PDF)
  

  This paper describes the three Fast Downward Aidos portfolios we submitted to the Unsolvability International Planning Competition 2016. All three Aidos variants are implemented in the Fast Downward planning system (Helmert2006). We use a pool of techniques as a basis for our portfolios, including various techniques already implemented Fast Downward, as well as three newly developed techniques to prove unsolvability. We used automatic algorithm configuration to find a good Fast Downward configuration for each of a set of test domains and used the resulting data to select the components, their order and their time slices for our three portfolios. For Aidos 1 and 2 we made this selection manually, resulting in two portfolios comprised mostly of the three new techniques. Aidos 1 distributes the 30 minutes based on our experiments, while Aidos 2 distributes the time uniformly. Aidos 3 contains unmodified configurations from the tuning process with time slices automatically optimized for the number of solved instances per time. It is based both on the new and existing Fast Downward components. The remainder of this planner abstract is organized as follows. First, we describe the three newly developed techniques. Second, we list the previously existing components of Fast Downward that we have used for configuration. Third, we describe the benchmarks used for training and test sets. Fourth, we describe the algorithm configuration process in more detail. Finally, we briefly describe the resulting portfolios.

• Robert C. Holte, Dr. Yusra Alkhazraji and Martin Wehrle.
  A Generalization of Sleep Sets Based on Operator Sequence Redundancy.
  In Proceedings of the 29th AAAI Conference on Artificial Intelligence (AAAI 2015). AAAI Press 2015.
  (Show abstract) (Hide abstract) (PDF)
  

  Pruning techniques have recently been shown to speed up search algorithms by reducing the branching factor of large search spaces. One such technique is sleep sets, which were originally introduced as a pruning technique for model checking, and which have recently been investigated on a theoretical level for planning. In this paper, we propose a generalization of sleep sets and prove its correctness. While the original sleep sets were based on the commutativity of operators, generalized sleep sets are based on a more general notion of operator sequence redundancy. As a result, our approach dominates the original sleep sets variant in terms of pruning power. On a practical level, our experimental evaluation shows the potential of sleep sets and their generalizations on a large and common set of planning benchmarks.

• Dr. Yusra Alkhazraji, Michael Katz, Robert Mattmüller, Florian Pommerening, Alexander Shleyfman and Martin Wehrle.
  Metis: Arming Fast Downward with Pruning and Incremental Computation (planner abstract).
  In the 8th International Planning Competition (IPC 2014) (deterministic track). 2014.
  (Show abstract) (Hide abstract) (PDF)
  

  Metis is a sequential optimal planner that implements three components on top of the Fast Downward planning system (Helmert 2006). The planner performs an A ∗ search using the following three major components: • an admissible incremental LM-cut heuristic (Pommeren- ing and Helmert 2013), • a symmetry based pruning technique (Domshlak, Katz, and Shleyfman 2012), and • a partial order reduction based pruning technique based on strong stubborn sets (Wehrle and Helmert 2012). Each of those techniques was extended to support conditional effects. In addition, Metis features a flexible invocation of partial order reduction based pruning. In what follows, we describe each of these components in detail.

• Martin Wehrle, Malte Helmert, Dr. Yusra Alkhazraji and Robert Mattmüller.
  The Relative Pruning Power of Strong Stubborn Sets and Expansion Core.
  In Proceedings of the 23rd International Conference on Automated Planning and Scheduling (ICAPS13). 2013.
  (Show abstract) (Hide abstract) (PDF)
  

  In the last years, pruning techniques based on partial order reduction have found increasing attention in the planning community. One recent result is that the expansion core method is a special case of the strong stubborn sets method proposed in model checking. However, it is still an open question if there exist efficiently computable strong stubborn sets with strictly higher pruning power than expansion core. In this paper, we prove that the pruning power of strong stubborn sets is strictly higher than the pruning power of expansion core even for a straight-forward instantiation of strong stubborn sets. This instantiation is as efficiently computable as expansion core. Hence, our theoretical results suggest that strong stubborn sets should be preferred to expansion core. Our empirical evaluation on all optimal benchmarks from the international planning competitions up to 2011 supports the theoretical results.

• Dr. Yusra Alkhazraji, Martin Wehrle, Robert Mattmüller and Malte Helmert.
  A Stubborn Set Algorithm for Optimal Planning.
  In Proceedings of the 20th European Conference on Artificial Intelligence (ECAI 2012). 2012.
  (Show abstract) (Hide abstract) (PDF)
  

  We adapt a partial order reduction technique based on stubborn sets, originally proposed for detecting dead ends in Petri Nets, to the setting of optimal planning. We demonstrate that stubborn sets can provide significant state space reductions on standard planning benchmarks, outperforming the expansion core method.

Christian Becker-Asano

• Nicolas Riesterer, Christian Becker-Asano, Julien Hué, Christian Dornhege and Bernhard Nebel.
  The Hybrid Agent MARCO.
  In Proceedings of the 16th International Conference on Multimodal Interaction, pp. 80-81. 2014.
  (Show abstract) (Hide abstract) (PDF)
  

  We present MARCO, a hybrid, chess playing agent equipped with a custom-built robotic arm and a virtual agent’s face displaying emotions. MARCO was built to investigate the hypothesis that hybrid agents capable of displaying emotions make playing chess more personal and enjoyable. In addition, we aim to explore means of achieving emotional contagion between man and machine.

• Christian Becker-Asano, Kai Oliver Arras and Bernhard Nebel.
  Robotic Tele-presence with DARYL in the Wild.
  In Proceedings of the 2nd International Confernce on Human-Agent Interaction, pp. 91-95. 2014.
  (Show abstract) (Hide abstract) (PDF)
  

  This paper describes the results of a qualitative analysis of questionnaire data collected during a public exhibition of our robotic tele-presence system. In Summer 2013 the mildly humanized robot DARYL could be tried out by the general public during our University’s science fair in the city center. People were given the chance to communicate through the robot with their peers and to perceive the world through the “eyes” and “ears” of the robot by means of a head-mounted display with attached headphones. An operator’s voice was instanta- neously transmitted to the robot’s location and his or her head movements were tracked to enable direct, intuitive control of the robot’s head movements. Twenty-seven people were interviewed in a structured way about their impressions and opinions after having either operated or interacted with the tele-operated robot. A careful analysis of the acquired data reveals a rather positive evaluation of the tele-presence system and interesting opinions about suitable application areas. These findings may guide designers of robotic tele-presence systems, a research area of increasing popularity.

• Christian Becker-Asano, Eduardo Meneses, Nicolas Riesterer, Julien Hué, Christian Dornhege and Bernhard Nebel.
  The Hybrid Agent MARCO: A Multimodal Autonomous Robotic Chess Opponent.
  In Proceedings of the 2nd International Confernce on Human-Agent Interaction, pp. 173-176. 2014.
  (Show abstract) (Hide abstract) (PDF)
  

  We present MARCO, a hybrid, chess playing agent equipped with a custom-built robotic arm and a virtual agent’s face displaying emotions. MARCO was built to investigate the hypothesis that hybrid agents capable of displaying emo- tions make playing chess more personal and enjoyable. In addition, we aim to explore means of achieving emotional contagion between man and machine.

• Felix Burkhardt, Christian Becker-Asano, Edmon Begoli, Roddy Cowie, Gerhard Fobe, Patrick Gebhard, Abe Kazemzadeh, Ingmar Steiner and Tim Llewellyn.
  Application of EmotionML.
  In 5th Intl. Workshop on Emotion, Social Signals, Sentiment & Linked Open Data (ES^3LOD), pp. 1-5. 2014.
  

• Christian Becker-Asano, Severin Gustorff, Kai Oliver Arras and Bernhard Nebel.
  On the Effect of Operator Modality on Social and Spatial Presence during Teleoperation of a Human-Like Robot.
  In Third Interantional Symposium on New Frontiers in Human-Robot Interaction at AISB50. 2014.
  (Show abstract) (Hide abstract) (PDF)
  

  With an increasing availability of affordable and effective robotic telepresence systems, key questions in the design of such systems arise, in particular when they aim at untrained users. Previous research regarding telepresence systems has focused on the (mobile) robotic platforms themselves or the differences between virtual as compared to physical representations thereof. The design space of the operator interface, i.e., the operator modality, with its potential impact on presence, however, has not been explored systematically. This paper reports results of an empirical study investigating how two different operator modalities impact the perceived spatial and social presence of operators in dyadic remote multi-modal interaction. The robot Daryl, used as telepresence medium in our study, features three degrees of freedom in its head unit as well as a stereo camera system. This enabled the transmission of a stereo, first-person perspective, which was used by the operator in combination with a head-mounted display whose movements were tracked to drive the robot’s head. Compared to a previously realized console-based operator interface, our results show significantly higher spatial as well as social presence for the head-mounted display modality while no significant difference in task performance was found. We conclude that for robotic telepresence platforms with mobile head units and stereo camera systems it seems advisable to use a head-mounted display as part of the teleoperation interface in order to provide operators with a particularly immersive remote presence experience.

• Christian Becker-Asano, Felix Ruzzoli, Christoph Hölscher and Bernhard Nebel.
  A Multi-Agent System based on Unity 4 for Virtual Perception and Wayfinding.
  Transportation Research Procedia 2, pp. 425-455. 2014.
  (Show abstract) (Hide abstract) (PDF)
  

  We developed a multi-agent system that is based on the game engine Unity 4 and allows simulating three-dimensional (3D) way- finding behavior of up to 600 airport passengers at a simulation rate of 60 Hz on an average gaming PC. Virtual 3D perception algorithms are implemented so that the agents dynamically check their respective surroundings for visible signs. Each sign is annotated with the direction of one or more exits and with meta-information such as its readability. Thus, based on findings derived from cognitive science experiments, the agents are modeled to sometimes misinterpret this information. Otherwise, they interpret the sign relative to its location and are then steered into the corresponding direction. This simulation framework was also combined with the head-mounted display “Oculus Rift” to let experiment participants find their way in the Virtual Reality environment.

• A. M. Rosenthal-von der Pütten, N. C. Krämer, Christian Becker-Asano, K. Ogawa, S. Nishio and H. Ishiguro.
  The uncanny in the wild. Analysis of unscripted human-android interaction in the field.
  Intl. Journal of Social Robotics 6 (1), pp. 67-83. 2014.
  

• Kiril Kiryazov, Robert Lowe, Christian Becker-Asano and Marco Randazzo.
  The Role of Arousal in Two-Resource Problem Tasks for Humanoid Service Robots.
  In Proc. IEEE Intl. Symposium on Robot and Human Interactive Communication (RO-MAN'13). 2013.
  

• Christian Becker-Asano, Dali Sun, Corinna N. Scheel, Brunna Tuschen-Caffier and Bernhard Nebel.
  Analyzing for emotional arousal in HMD-based head movements during a virtual emergency.
  In Intl. Workshop on Emotion and Computing in conj. with KI2013. 2013.
  (Show abstract) (Hide abstract) (PDF)
  

  This paper reports on results of a statistical analysis of human players' head-movements. Forty-one participants were asked to cope with an unexpected emergency in a virtual parking lot. Before the virtual reality exposure began, half of the participants watched an emotion-inducing movie clip and the other half an emotionally neutral one. The analysis of the acquired questionnaire data reveals, however, that this emotion induction method seems to have been rather ineffective. Thus, it is not surprising that only very weak between group effects are found when analyzing for differences in head movements around the emergency event. In general, horizontal head movement speed is found to be on average significantly faster during the first fifteen seconds directly after the emergency event as compared to just before and another fifteen seconds later. These findings are in line with previous results of an analysis of the acquired physiological data, further substantiating the conclusions drawn.

• Christian Becker-Asano, Philip Stahl, Marco Ragni, Jean-Claude Martin, Matthieu Courgeon and Bernhard Nebel.
  An affective virtual agent providing embodied feedback in the paired associate task: system design and evaluation.
  In Proc. of the 13th. Intl. Conf. on Intelligent Virtual Agents (IVA 2013), pp. 406-415. 2013.
  (Show abstract) (Hide abstract) (PDF)
  

  An affective, virtual agent is presented that acts as a teacher in the classical paired associate task. It is explained, why and how the virtual agent framework MARC was combined with the cognitive architecture ACT-R, the affect simulation architecture WASABI, and the voice-synthesis module OpenMARY. The agent's affective feedback capabilities are evaluated through an empirical study, in which participants had to solve association tasks. We expected that (1) the presentation of the task by a (neutral) virtual agent would change a learner's performance and that (2) the additional simulation and expression of emotions would impact a learner's performance as well. Finally, we discuss reasons for the lack of statistically significant differences as well as planned future application scenarios of our affective agent framework.

• Christian Becker-Asano, Severin Gustorff, Kai Oliver Arras, Kohei Ogawa, Shuichi Nishio, Hiroshi Ishiguro and Bernhard Nebel.
  Robot embodiment, operator modality, and social interaction in tele-existence: a project outline.
  In Proceedings of the 8th ACM/IEEE international conference on Human-robot interaction, pp. 79-80. 2013.
  (Show abstract) (Hide abstract) (PDF)
  

  This paper outlines our ongoing project, which aims to investigate the effects of robot embodiment and operator modality on an operator’s task efficiency and concomitant level of copresence in remote social interaction. After a brief introduction to related work has been given, five research questions are presented. We discuss how these relate to our choice of the two robotic embodiments “DARYL” and “Geminoid F” and the two operator modalities “console interface” and “head-mounted display”. Finally, we postulate that the usefulness of one operator modality over the other will depend on the type of situation an operator has to deal with. This hypothesis is currently being investigated empirically using DARYL at Freiburg University.

• Nicole C. Krämer, Stefan Kopp, Christian Becker-Asano and Nicole Sommer.
  Smile and the world will smile with you-The effects of a virtual agent's smile on users’ evaluation and behavior.
  International Journal of Human-Computer Studies 71 (3), pp. 335-349. 2013.
  

• Christian Becker-Asano and Hiroshi Ishiguro.
  Intercultural Differences in Decoding Facial Expressions of the Android Robot Geminoid F.
  Journal of Artificial Intelligence and Soft Computing Research 1 (3), pp. 215-231. 2012.
  (Show abstract) (Hide abstract)
  

  As android robots become increasingly sophisticated in their technical as well as artistic design, their non-verbal expressiveness is getting closer to that of real humans. Accordingly, this paper presents results of two online surveys designed to evaluate a female android's facial display of five basic emotions. Being interested in intercultural differences we prepared both surveys in English, German, as well as Japanese language, and we not only found that in general our design of the emotional expressions "fearful" and "surprised" (...)

• Christian Becker-Asano, Kai Oliver Arras, Bernhard Nebel and Hiroshi Ishiguro.
  The Effect of Anthropomorphism on Social Tele-Embodiment.
  In IROS 2012 Workshop on Human-Agent Interaction. 2012.
  (Show abstract) (Hide abstract) (PDF)
  

  This paper outlines our approach to explore the impact of using two different robotic embodiments on an oper- ators ability to convey emotional and conversational nonverbal signals to a distant interlocutor. Although a human’s ability to produce and interpret complex, dynamic facial expressions is seen as an important factor for human-human social interac- tion, it remains controversial in humanoid/android robotics, whether recreating such expressiveness is really worth the technical challenge, or not. In fact, one way to avoid the risk of giving rise to uncanny feelings in human observers is to follow an abstract design for humanoid robots. This question is also relevant in the context of mediated interaction using tele-operation technology, as soon as robotic embodiments are involved. Thus, this paper presents our current project, in which we are comparing the efficiency of transmitting nonverbal signals by means of “Daryl” featuring an abstract, mildly humanized design, against that of “Geminoid F”, which features a highly anthropomorphic design. The ability of both of these robots to convey emotions by means of body movements has been successfully evaluated before, but using this ability to transmit nonverbal signals during remote conversation and comparing the resp. efficiencies has not yet been done.

• Birgit Kleim, Thomas Ehrig, Corinna Scheel, Christian Becker-Asano, Bernhard Nebel and Brunna Tuschen-Caffier.
  Bewältigungsverhalten in Notfallsituationen aus klinisch-psychologischer Perspektive.
  Zeitschrift für Klinische Psychologie und Psychotherapie 41 (3), pp. 166-179. 2012.
  (Show abstract) (Hide abstract) (Online; DOI)
  

  Ziel des vorliegenden Beitrags ist es, eine aktuelle Übersicht zu Annahmen und Befunden zu geben, die Hinweise darauf geben, welche Reaktionen bzw. welches Verhalten für die Bewältigung von Notfällen oder traumatischen Erlebnissen hilfreich bzw. gesundheitsförderlich sind. Ließen sich konkrete Aspekte von Bewältigungsverhalten während traumatischer Situationen identifizieren, die besonders adaptiv in Bezug auf die psychische bzw. psychobiologische Anpassung sind, so könnte dieses Wissen perspektivisch zur Entwicklung von Präventions- und Trainingsmaßnahmen genutzt werden. Der Beitrag beschreibt einleitend Traumareaktionen, psychische Traumafolgestörungen und deren Prävalenzraten und gibt eine knappe Übersicht über Prädiktoren für psychische Störungen in Folge traumatischer Erlebnisse. Im Unterschied zu dem Beitrag von Becker-Nehring, Witschen und Bengel (in diesem Heft) fokussiert unser Beitrag vor allem auf die Posttraumatische Belastungsstörung als Traumafolgestörung und auf Bewältigungsverhalten während einer Notfallsituation. Bewältigungsverhalten während und nach einer traumatischen Situation kann zum Teil auch im Forschungslabor experimentell untersucht werden, indem z. B. Methoden der Virtuellen Realität genutzt werden. Dies ist ein weiterer Fokus des Beitrags.

• Corinna N. Scheel, Birgit Kleim, Julian Schmitz, Christian Becker-Asano, Dali Sun, Bernhard Nebel and Brunna Tuschen-Caffier.
  Psychophysiologische Belastungsreaktivität nach einem simulierten Feuer in einer Parkgarage.
  Zeitschrift für Klinische Psychologie und Psychotherapie 41 (3), pp. 180-189. 2012.
  (Show abstract) (Hide abstract) (Online; DOI)
  

  Theoretischer Hindergrund: Bewältigungsverhalten in Notfallsituationen wird meistens retrospektiv erfasst oder ist aufgrund der Verschiedenheit der Notfallsituationen schlecht vergleichbar. Methoden der Virtuellen Realität (VR) ermöglichen die Erfassung von Verhaltensparametern und psychophysiologischen Belastungsreaktionen während eines belastenden Ereignisses und erlauben zudem das standardisierte Wiederholen für mehrere Personen. Fragestellung: Ziel unserer Studie war es, ein neues Notfallszenario (Feuer in einer Parkgarage) in VR zu entwickeln und zu testen, ob sich anhand dessen substanzielle psychische und physiologische Belastungsreaktionen induzieren lassen. Methode: Mehrfach im Untersuchungsablauf wurden das emotionale Erleben und physiologische Parameter erhoben. Ergebnisse: Das VR Szenario führte bei den teilnehmenden Probanden sowohl zu subjektiven als auch zu physiologischen Veränderungen im Sinne einer Stressinduktion. Das von uns entwickelte Szenario erscheint daher brauchbar, Verhaltensstrategien und Bewältigungsverhalten in Notfallsituationen zu simulieren. Schlussfolgerungen: Möglichkeiten und Grenzen der VR-Methode mit Blick auf klinisch-psychologische Implikationen werden diskutiert.

• Stephanie Embgen, Matthias Luber, Christian Becker-Asano, Marco Ragni, Vanessa Evers and Kai Oliver Arras.
  Robot-Specific Social Cues in Emotional Body Language.
  In Proc. IEEE International Symposium on Robot and Human Interactive Communication (RO-MAN'12), pp. 1019-1025. 2012.
  (Show abstract) (Hide abstract)
  

  Humans use very sophisticated ways of bodily emotion expression combining facial expressions, sound, gestures and full body posture. Like others, we want to apply these aspects of human communication to ease the interaction between robots and users. In doing so we believe there is a need to consider what abstraction of human social communicative behaviors is appropriate for robots. The study reported in this paper is a pilot study to not offer simulated emotion but to offer an abstracted robot version of emotion expressions (...)

• Christian Becker-Asano.
  Affective Computing Combined with Android Science.
  KI - Künstliche Intelligenz Vol. 25, pp. 245-250. 2011.
  (PDF) (BIB)
  

• Christian Becker-Asano, Dali Sun, Birgit Kleim, Corinna Scheel, Brunna Tuschen-Caffier and Bernhard Nebel.
  Outline of an Empirical Study on the Effects of Emotions on Strategic Behavior in Virtual Emergencies.
  In Affective Computing and Intelligent Interaction, pp. 508-517. 2011.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  The applicability of appropriate coping strategies is important in emergencies or traumatic experiences such as car accidents or human violence. In this context, emotion regulation and decision making are relevant. However, research on human reactions to traumatic experiences is very challenging and most existing research uses retrospective assessments of these variables of interest. Thus, we are currently developing and evaluating novel methods to investigate human behavior in cases of emergency. Virtual reality scenarios of emergencies are employed to enable an immersive interactive engagement (e.g., dealing with fire inside a building) based on the modification of Valve’s popular Source 2007 game engine. This paper presents our ongoing research project, which aims at the empirical investigation of human strategic behavior under the influence of emotions while having to cope with virtual emergencies.

• Christian Becker-Asano.
  Invited Commentary: On Guiding the Design of an Ill-defined Phenomenon.
  International Journal of Synthetic Emotions Vol. 2 (2), pp. 66-67. 2011.
  (PDF) (BIB)
  

• Brunna Tuschen-Caffier, Birgit Kleim, Christian Becker-Asano, Dali Sun, Bernhard Nebel and Corinna Scheel.
  Bewältigungsverhalten in virtuellen Notfallsituationen.
  In 7. Workshop Kongress für Psychologie und Psychotherapie. 2011.
  (BIB)
  

• Christian Becker-Asano, Dali Sun, Birgit Kleim, Corinna N. Scheel, Brunna Tuschen-Caffier and Bernhard Nebel.
  CoVE: Coping in Virtual Emergencies.
  In Workshop on Emotion and Computing - Current Research and Future Impact, p. 1. 2011.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  The applicability of appropriate coping strategies is important in emergencies or traumatic experiences such as car accidents or human violence. However, research on human reactions to traumatic experiences is very challenging and most existing research uses retrospective assessments of these variables of interest. Thus, we are currently developing and evaluating novel methods to investigate human behavior in cases of emergency. Virtual Reality (VR) scenarios of emergencies are employed to enable an immersive interactive engagement (e.g., dealing with fire inside a building) based on the modification of Valve’s popular Source 2007 game engine. Preliminary results of a first empirical study (cp. Figure 1) suggest that our VR scenario has a similar fear-inducing effect as a short movie clip (Becker- Asano, Sun, Kleim, Scheel, Tuschen-Caffier, and Nebel, 2011), which previously has been evaluated to induce fear. In addition, the neutral VR experiences during the training sessions did never elicit fear in our participants, letting us conclude that the interactively presented emergency itself was indeed the fear eliciting factor in the experimental sessions. In the long run, we aim at a more detailed analysis that includes the personality questionnaire and physiological data, which will be analyzed in correlation with the trajectories of the participants in the VR emergency.

Gregor Behnke

• Roman Barták, Simona Ondrčková, Gregor Behnke and Pascal Bercher.
  Correcting Hierarchical Plans by Action Deletion.
  In Proceedings of the 18th International Conference on Principles of Knowledge Representation and Reasoning (KR-2021). 2021.
  (PDF)
  

• Roman Barták, Simona Ondrčková, Gregor Behnke and Pascal Bercher.
  Correcting Hierarchical Plans by Action Deletion.
  In Proceedings of the Fourth ICAPS Workshop on Hierarchical Planning. 2021.
  (PDF)
  

• Roman Barták, Simona Ondrčková, Gregor Behnke and Pascal Bercher.
  On the Verification of Totally-Ordered HTN Plans.
  In Proceedings of the Fourth ICAPS Workshop on Hierarchical Planning, pp. 44-48. 2021.
  (PDF)
  

• Daniel Höller, Julia Whichlacz, Pascal Bercher and Gregor Behnke.
  Compiling HTN Plan Verification Problems into HTN Planning Problems.
  In Proceedings of the Fourth ICAPS Workshop on Hierarchical Planning, pp. 8-15. 2021.
  (PDF)
  

• Pascal Bercher, Gregor Behnke, Matthias Kraus, Marvin Schiller, Dietrich Manstetten, Michael Dambier, Michael Dorna, Wolfgang Minker, Birte Glimm and Susanne Biundo.
  Do It Yourself, but Not Alone: Companion-Technology for Home Improvement - Bringing a Planning-Based Interactive DIY Assistant to Life.
  Künstliche Intelligenz -- Special Issue on NLP and Semantics. 2021.
  (Show abstract) (Hide abstract)
  

  We report on the technology transfer project "Do it yourself, but not alone: Companion-Technology for Home Improvement" that was carried out by Ulm University in cooperation with Robert Bosch GmbH. We developed a prototypical assistance system that assists a Do It Yourself (DIY) handyman in carrying out DIY projects. The assistant, based on various AI and dialog management capabilities, generates a sequence of detailed instructions that users may just follow or adapt according to their individual preferences. It features explanation capabilities as well as pro-active support based on communication with the user as well as with the involved tools. We report on the project’s main achievements, including the findings of various empirical studies conducted in various development stages of the prototype.

• Gregor Behnke.
  Block Compression and Invariant Pruning for SAT-based Totally-Ordered HTN Planning.
  In Proceedings of the 31st International Conference on Automated Planning and Scheduling (ICAPS 2021), pp. 25-35. 2021.
  (Show abstract) (Hide abstract) (PDF)
  

  Translations into propositional logic are currently one of the most efficient techniques for solving Totally-Ordered HTN planning problems. The current encodings iterate over the maximum allowed depth of decomposition. Given this depth, they compute a tree that represents all possible decompositions up to this depth. Based on this tree, a formula in propositional logic is created. We show that much of the computed tree is actually useless as it cannot possibly belong to a solution. We provide a technique for removing (parts of) these useless structures using state invariants. We further show that is often not necessary to encode all leafs of this tree as separate timesteps, as the prior encodings did. Instead, we can compress the leafs into blocks and encode all leafs of a block as one timestep. We show that these changes provide an improvement over the state-of-the-art in HTN planning.

• Daniel Höller and Gregor Behnke.
  Loop Detection in the PANDA Planning System.
  In Proceedings of the 31st International Conference on Automated Planning and Scheduling (ICAPS 2021), pp. 168-173. 2021.
  (Show abstract) (Hide abstract) (PDF)
  

  The International Planning Competition (IPC) in 2020 was the first one for a long time to host tracks on Hierarchical Task Network (HTN) planning. HyperTensioN, the winner of the tack on totally-ordered problems, comes with an interesting technique: it stores parts of the decomposition path in the state to mark expanded tasks and forces its depth first search to leave recursive structures in the hierarchy. This can be seen as a form of loop detection (LD) -- a technique that is not very common in HTN planning. This might be due to the spirit of encoding enough advice in the model to find plans (so that loop detection is simply not necessary), or because it becomes a computationally hard task in the general (i.e. partially-ordered) setting. We integrated several approximate and exact techniques for LD into the progression search of the HTN planner PANDA. We test our techniques on the benchmark set of the IPC 2020. Both in the partial ordered and total ordered track, PANDA with LD performs better than the respective winner of the competition.

• Daniel Höller, Gregor Behnke, Pascal Bercher and Susanne Biundo.
  The PANDA Framework for Hierarchical Planning.
  KI - Künstliche Intelligenz. 2021.
  (Online)
  

• Gregor Behnke and David Speck.
  Symbolic Search for Optimal Total-Order HTN Planning.
  In Proceedings of the 35th AAAI Conference on Artificial Intelligence (AAAI 2021), p. 11744–11754. 2021.
  (Show abstract) (Hide abstract) (PDF)
  

  Symbolic search has proven to be a useful approach to optimal classical planning. In Hierarchical Task Network (HTN) planning, however, there is little work on optimal planning. One reason for this is that in HTN planning, most algorithms are based on heuristic search, and admissible heuristics have to incorporate the structure of the task network in order to be informative. In this paper, we present a novel approach to optimal (totally-ordered) HTN planning, which is based on symbolic search. An empirical analysis shows that our symbolic approach outperforms the current state of the art for optimal totally-ordered HTN planning.

• Matthias Kraus, Marvin Schiller, Gregor Behnke, Pascal Bercher, Michael Dorna, Michael Dambier, Birte Glimm, Susanne Biundo and Wolfgang Minker.
  ``Was that successful?'' On Integrating Proactive Meta-Dialogue in a DIY-Assistant using Multimodal Cues.
  In Proceedings of the 2020 International Conference on Multimodal Interaction (ICMI 2020). 2020.
  (PDF)
  

• Daniel Höller, Pascal Bercher, Gregor Behnke and Susanne Biundo.
  HTN Plan Repair via Model Transformation.
  In Proceedings of the 42+1st Annual German Conference on Artificial Intelligence (KI 2020). 2020.
  (PDF)
  

• Daniel Höller, Pascal Bercher and Gregor Behnke.
  Delete- and Ordering-Relaxation Heuristics for HTN Planning.
  In Proceedings of International Joint Conference on Artificial Intelligence (IJCAI 2020). 2020.
  

• Roman Bartak, Simona Ondrckova, Adrien Maillard, Gregor Behnke and Pascal Bercher.
  A Novel Parsing-based Approach for Verification of Hierarchical Plans.
  In Proceedings of the 32nd International Conference on Tools with Artificial Intelligence (ICTAI 2020). 2020.
  

• Daniel Höller, Pascal Bercher, Gregor Behnke and Susanne Biundo.
  HTN Planning as Heuristic Progression Search.
  Journal of Artificial Intelligence Research, pp. 835-880. 2020.
  

• Gregor Behnke, Pascal Bercher, Matthias Kraus, Marvin Schiller, Kristof Mickeleit, Häge Timo, Michael Dorna, Michael Dambier, Wolfgang Minker, Birte Glimm and Susanne Biundo.
  New Developments for Robert - Assisting Novice Users Even Better in DIY Projects.
  In Proceedings of the 30th International Conference on Automated Planning and Scheduling (ICAPS 2020), pp. 343-347. 2020.
  (PDF)
  

• Daniel Höller, Gregor Behnke, Pascal Bercher, Susanne Biundo, Humbert Fiorino, Damien Pellier and Ron Alford.
  HDDL - A Language to Describe Hierarchical Planning Problems.
  In Proceedings of the 34th AAAI Conference on Artificial Intelligence (AAAI 2020), pp. 6-17. 2020.
  (PDF)
  

• Gregor Behnke, Daniel Höller, Alexander Schmid, Pascal Bercher and Susanne Biundo.
  On Succinct Groundings of HTN Planning Problems.
  In Proceedings of the 34th AAAI Conference on Artificial Intelligence (AAAI 2020), pp. 9775-9784. 2020.
  (Show abstract) (Hide abstract) (PDF)
  

  Both search-based and translation-based planning systemsusually operate on grounded representations of the problem. Planning models, however, are commonly defined using lifted description languages. Thus, planning systems usually generate a grounded representation of the lifted model as a pre-processing step. For HTN planning models, only one method to ground lifted models has been published so far. In this paper we present a new approach for grounding HTN planning problems that produces smaller groundings in a shorter timespan than the previously published method.

• Gregor Behnke, Daniel Höller and Susanne Biundo.
  Finding Optimal Solutions in HTN Planning - A SAT-based Approach.
  In Proceedings of the 28th International Joint Conference on Artificial Intelligence (IJCAI 2019), pp. 5500-5508. 2019.
  (PDF)
  

• Gregor Behnke, Marvin Schiller, Matthias Kraus, Pascal Bercher, Mario Schmautz, Michael Dorna, Michael Dambier, Wolfgang Minker, Birte Glimm and Susanne Biundo.
  Alice in DIY-Wonderland or: Instructing novice users on how to use tools in DIY projects.
  AI Communications, pp. 31-57. 2019.
  (PDF)
  

• Gregor Behnke, Daniel Höller and Susanne Biundo.
  Bringing Order to Chaos - A Compact Representation of Partial Order in SAT-based HTN Planning.
  In Proceedings of the 33rd AAAI Conference on Artificial Intelligence (AAAI 2019), pp. 7520-7529. 2019.
  (PDF)
  

• Daniel Höller, Pascal Bercher, Gregor Behnke and Susanne Biundo.
  On Guiding Search in HTN Planning with Classical Planning Heuristics.
  In Proceedings of the 28th International Joint Conference on Artificial Intelligence (IJCAI 2019). 2019.
  (PDF)
  

• Gregor Behnke, Daniel Höller, Pascal Bercher, Susanne Biundo, Humbert Fiorino, Damien Pellier and Ron Alford.
  Hierarchical Planning in the IPC.
  In Proceedings of 2019 Workshop on the International Planning Competition (WIPC 2019), pp. 40-47. 2019.
  (PDF)
  

• Daniel Höller, Gregor Behnke, Pascal Bercher, Susanne Biundo, Humbert Fiorino, Damien Pellier and Ron Alford.
  HDDL - A Language to Describe Hierarchical Planning Problems.
  In Proceedings of the Second ICAPS Workshop on Hierarchical Planning, pp. 6-17. 2019.
  (PDF)
  

• Gregor Behnke, Daniel Höller, Pascal Bercher and Susanne Biundo.
  More Succinct Grounding of HTN Planning Problems - Preliminary Results.
  In Proceedings of the Second ICAPS Workshop on Hierarchical Planning, pp. 40-48. 2019.
  (PDF)
  

• Matthias Kraus, Gregor Behnke, Pascal Bercher, Marvin Schiller, Susanne Biundo, Birte Glimm and Wolfgang Minker.
  A Multimodal Dialogue Framework for Cloud-Based Companion Systems.
  In Proceedings of the 9th International Workshop on Spoken Dialogue Systems Technology (IWSDS 2018), pp. 405-410. 2018.
  (PDF)
  

• Marvin Schiller, Gregor Behnke, Pascal Bercher, Matthias Kraus, Michael Dorna, Felix Richter, Susanne Biundo, Birte Glimm and Wolfgang Minker.
  Evaluating Knowledge-Based Assistance for DIY.
  In Proceedings of MCI Workshop ``Digital Companion'', pp. 925-930. 2018.
  (PDF)
  

• Gregor Behnke, Daniel Höller and Susanne Biundo.
  Tracking Branches in Trees - A Propositional Encoding for solving Partially-Ordered HTN Planning Problems.
  In Proceedings of the 30th International Conference on Tools with Artificial Intelligence (ICTAI 2018), pp. 73-80. 2018.
  (PDF)
  

• Gregor Behnke and Susanne Biundo.
  X and more Parallelism: Integrating LTL-Next into SAT-based Planning with Trajectory Constraints While Allowing for Even More Parallelism.
  Inteligencia Artificial 21(62), pp. 75-90. 2018.
  (PDF)
  

• Gregor Behnke, Daniel Höller and Susanne Biundo.
  totSAT - Totally-Ordered Hierarchical Planning through SAT.
  In Proceedings of the 32nd AAAI Conference on Artificial Intelligence (AAAI 2018), pp. 6110-6118. 2018.
  (PDF)
  

• Gregor Behnke, Marvin Schiller, Matthias Kraus, Pascal Bercher, Mario Schmautz, Michael Dorna, Wolfgang Minker, Birte Glimm and Susanne Biundo.
  Instructing Novice Users on How to Use Tools in DIY Projects.
  In Proceedings of the 27th International Joint Conference on Artificial Intelligence and the 23rd European Conference on Artificial Intelligence (IJCAI-ECAI 2018), pp. 5805-5807. 2018.
  (PDF)
  

• Gregor Behnke, Daniel Höller and Susanne Biundo.
  Tracking Branches in Trees - A Propositional Encoding for solving Partially-Ordered HTN Planning Problems.
  In Proceedings of the First ICAPS Workshop on Hierarchical Planning, pp. 40-47. 2018.
  (PDF)
  

• Daniel Höller, Pascal Bercher, Gregor Behnke and Susanne Biundo.
  HTN Plan Repair Using Unmodified Planning Systems.
  In Proceedings of the First ICAPS Workshop on Hierarchical Planning, pp. 26-30. 2018.
  (PDF)
  

• Gregor Behnke and Susanne Biundo.
  X and more Parallelism - Integrating LTL-Next into SAT-based Planning with Trajectory Constraints while Allowing for even more Parallelism.
  In Proceedings of the Workshop on Constraint Satisfaction Techniques for Planning and Scheduling Problems (COPLAS 2018), pp. 1-10. 2018.
  (PDF)
  

• Daniel Höller, Gregor Behnke, Pascal Bercher and Susanne Biundo.
  Plan and Goal Recognition as HTN Planning.
  In Proceedings of the 30th International Conference on Tools with Artificial Intelligence (ICTAI 2018), pp. 466-473. 2018.
  (PDF)
  

• Daniel Höller, Pascal Bercher, Gregor Behnke and Susanne Biundo.
  A Generic Method to Guide HTN Progression Search with Classical Heuristics.
  In Proceedings of the 28th International Conference on Automated Planning and Scheduling (ICAPS 2018), pp. 114-122. 2018.
  (PDF)
  

• Daniel Höller, Pascal Bercher, Gregor Behnke and Susanne Biundo.
  Plan and Goal Recognition as HTN Planning.
  In Proceedings of the AAAI 2018 Workshop on Activity Plan and Intent Recognition (PAIR 2018), pp. 607-613. 2018.
  (PDF)
  

• Benedikt Leichtmann, Pascal Bercher, Daniel Höller, Gregor Behnke, Susanne Biundo, Verena Nitsch and Martin Baumann.
  Towards a Companion System Incorporating Human Planning Behavior - A Qualitative Analysis of Human Strategies.
  In Proceedings of the 3rd Transdisciplinary Conference on Support Technologies (TCST 2018), pp. 89-98. 2018.
  (PDF)
  

• Gregor Behnke, Daniel Höller and Susanne Biundo.
  This is a solution! (... but is it though?) - Verifying solutions of hierarchical planning problems.
  In Proceedings of the 27th International Conference on Automated Planning and Scheduling (ICAPS 2017), pp. 20-28. 2017.
  (PDF)
  

• Florian Nothdurft, Pascal Bercher, Gregor Behnke and Wolfgang Minker.
  User Involvement in Collaborative Decision-Making Dialog Systems.
  In Dialogues with Social Robots: Analyses Enablements and Evaluation, pp. 129-141. 2017.
  (PDF)
  

• Pascal Bercher, Gregor Behnke, Daniel Höller and Susanne Biundo.
  An Admissible HTN Planning Heuristic.
  In Proceedings of the 26th International Joint Conference on Artificial Intelligence (IJCAI 2017), pp. 480-488. 2017.
  (PDF)
  

• Marvin Schiller, Gregor Behnke, Mario Schmautz, Pascal Bercher, Matthias Kraus, Michael Dorna, Wolfgang Minker, Birte Glimm and Susanne Biundo.
  A Paradigm for Coupling Procedural and Conceptual Knowledge in Companion Systems.
  In Proceedings of the 2nd International Conference on Companion Technology (ICCT 2017). 2017.
  (PDF)
  

• Gregor Behnke, Benedikt Leichtmann, Pascal Bercher, Daniel Höller, Verena Nitsch, Martin Baumann and Susanne Biundo.
  Help me make a dinner! Challenges when assisting humans in action planning.
  In Proceedings of the 2nd International Conference on Companion Technology (ICCT 2017). 2017.
  (PDF)
  

• Gregor Behnke, Florian Nielsen, Marvin Schiller, Pascal Bercher, Matthias Kraus, Birte Glimm, Wolfgang Minker and Susanne Biundo.
  SLOTH - the Interactive Workout Planner.
  In Proceedings of the 2nd International Conference on Companion Technology (ICCT 2017). 2017.
  (PDF)
  

• Pascal Bercher, Daniel Höller, Gregor Behnke and Susanne Biundo.
  User-Centered Planning.
  In Companion Technology - A Paradigm Shift in Human-Technology Interaction, pp. 79-100. 2017.
  (PDF)
  

• Pascal Bercher, Felix Richter, Thilo Hörnle, Thomas Geier, Daniel Höller, Gregor Behnke, Florian Nielsen, Frank Honold, Schüssel Felix, Stephan Reuter, Wolfgang Minker, Michael Weber, Klaus Dietmayer and Susanne Biundo.
  Advanced User Assistance for Setting Up a Home Theater.
  In Companion Technology - A Paradigm Shift in Human-Technology Interaction, pp. 485-491. 2017.
  (PDF)
  

• Gregor Behnke, Florian Nielsen, Marvin Schiller, Denis Ponomaryov, Pascal Bercher, Birte Glimm, Wolfgang Minker and Susanne Biundo.
  To Plan for the User Is to Plan With the User - Integrating User Interaction Into the Planning Process.
  In Companion Technology - A Paradigm Shift in Human-Technology Interaction, pp. 123-144. 2017.
  (PDF)
  

• Pascal Bercher, Felix Richter, Frank Honold, Florian Nielsen, Felix Schüssel, Thomas Geier, Thilo Hörnle, Stephan Reuter, Daniel Höller, Gregor Behnke, Klaus Dietmayer, Wolfgang Minker, Michael Weber and Susanne Biundo.
  A Companion-System Architecture for Realizing Individualized and Situation-Adaptive User Assistance.
  In Technical Report - Ulm University. 2018.
  (PDF)
  

• Gregor Behnke, Daniel Höller, Pascal Bercher and Susanne Biundo.
  Change the Plan - How hard can that be?
  In Proceedings of the 26th International Conference on Automated Planning and Scheduling (ICAPS 2016), pp. 38-46. 2016.
  (PDF)
  

• Daniel Höller, Gregor Behnke, Pascal Bercher and Susanne Biundo.
  Assessing the Expressivity of Planning Formalisms through the Comparison to Formal Languages.
  In Proceedings of the 26th International Conference on Automated Planning and Scheduling (ICAPS 2016), pp. 158-165. 2016.
  (PDF)
  

• Ron Alford, Gregor Behnke, Daniel Höller, Pascal Bercher, Susanne Biundo and David W. Aha.
  Bound to Plan: Exploiting Classical Heuristics via Automatic Translations of Tail-Recursive HTN Problems.
  In Proceedings of the 26th International Conference on Automated Planning and Scheduling (ICAPS 2016), pp. 20-28. 2016.
  (PDF)
  

• Pascal Bercher, Daniel Höller, Gregor Behnke and Susanne Biundo.
  More than a Name? On Implications of Preconditions and Effects of Compound HTN Planning Tasks.
  In Proceedings of the 22nd European Conference on Artificial Intelligence (ECAI 2016), pp. 225-233. 2016.
  (PDF)
  

• Gregor Behnke, Daniel Höller and Susanne Biundo.
  On the Complexity of HTN Plan Verification and Its Implications for Plan Recognition.
  In Proceedings of the 25th International Conference on Automated Planning and Scheduling (ICAPS 2015), pp. 25-33. 2015.
  (PDF)
  

• Gregor Behnke, Denis Ponomaryov, Marvin Schiller, Ber\-cher Pascal, Florian Nothdurft, Birte Glimm and Susanne Biundo.
  Coherence Across Components in Cognitive Systems - One Ontology to Rule Them All.
  In Proceedings of the 24th International Joint Conference on Artificial Intelligence (IJCAI 2015), pp. 1442-1449. 2015.
  (PDF)
  

• Florian Nothdurft, Gregor Behnke, Pascal Bercher, Susanne Biundo and Wolfgang Minker.
  The Interplay of User-Centered Dialog Systems and AI Planning.
  In Proceedings of the 16th Annual Meeting of the Special Interest Group on Discourse and Dialogue (SIGDIAL 2015), pp. 344-353. 2015.
  (PDF)
  

• Pascal Bercher, Felix Richter, Thilo Hörnle, Thomas Geier, Daniel Höller, Gregor Behnke, Florian Nothdurft, Frank Honold, Wolfgang Minker, Michael Weber and Susanne Biundo.
  A Planning-based Assistance System for Setting Up a Home Theater.
  In Proceedings of the 29th AAAI Conference on Artificial Intelligence (AAAI 2015), pp. 4264-4265. 2015.
  (PDF)
  

• Gregor Behnke, Pascal Bercher, Susanne Biundo, Birte Glimm, Denis Ponomaryov and Marvin Schiller.
  Integrating Ontologies and Planning for Cognitive Systems.
  In Proceedings of the 28th International Workshop on Description Logics (DL 2015), pp. 338-360. 2015.
  (PDF)
  

• Pascal Bercher, Daniel Höller, Gregor Behnke and Susanne Biundo.
  User-Centered Planning - A Discussion on Planning in the Presence of Human Users.
  In Proceedings of the First International Symposium on Companion Technology (ISCT 2015), pp. 79-82. 2015.
  (PDF)
  

• Gregor Behnke, Marvin Schiller, Denis Ponomaryov, Florian Nothdurft, Ber\-cher Pascal, Wolfgang Minker, Birte Glimm and Susanne Biundo.
  A Unified Knowledge Base for Companion-Systems - A Case Study in Mixed-Initiative Planning.
  In Proceedings of the First International Symposium on Companion Technology (ISCT 2015), pp. 43-48. 2015.
  (PDF)
  

• Daniel Höller, Gregor Behnke, Pascal Bercher and Susanne Biundo.
  Language Classification of Hierarchical Planning Problems.
  In Proceedings of the 21st European Conference on Artificial Intelligence (ECAI 2014), pp. 447-452. 2014.
  (PDF)
  

Rolf-David Bergdoll

• Pascal Bachor, Rolf-David Bergdoll and Bernhard Nebel.
  The Multi-Agent Transportation Problem.
  In Proceedings of the 37th AAAI Conference on Artificial Intelligence (AAAI 2023), pp. 11525-11532. 2023.
  (Show abstract) (Hide abstract) (Online; DOI)
  

  We introduce the multi-agent transportation (MAT) problem, where agents have to transport containers from their starting positions to their designated goal positions. Movement takes place in a common environment where collisions between agents and between containers must be avoided. In contrast to other frameworks such as MAPF or MAPD, the agents are allowed to separate from the containers at any time, which allows for plans in scenarios that might otherwise be unsolvable and has the potential to reduce the makespan. We present a complexity analysis establishing NP-completeness and show how the problem can be reduced to a sequence of SAT problems when optimizing for makespan. Finally, the implementation is empirically evaluated relative to input characteristics, and it is compared to some variants of the MAT problem and a CBS-based MAPD implementation.

• Marius Lindauer, Rolf-David Bergdoll and Frank Hutter.
  An Empirical Study of Per-instance Algorithm Scheduling.
  In Proceedings of the 10th Learning and Intelligent Optimization Conference (LION 10), pp. 253-259. 2016.
  (Show abstract) (Hide abstract) (PDF)
  

  Algorithm selection is a prominent approach to improve a system’s performance by selecting a well-performing algorithm from a portfolio for an instance at hand. One extension of the traditional algorithm selection problem is to not only select one single algorithm but a schedule of algorithms to increase robustness. Some approaches exist for solving this problem of selecting schedules on a per-instance basis (e.g., the Sunny and 3S systems), but to date, a fair and thorough comparison of these is missing. In this work, we implement Sunny’s approach and dynamic schedules inspired by 3S in the flexible algorithm selection framework flexfolio to use the same code base for a fair comparison. Based on the algorithm selection library (ASlib), we perform the first thorough empirical study on the strengths and weaknesses of per-instance algorithm schedules. We observe that on some domains it is crucial to use a training phase to limit the maximal size of schedules and to select the optimal neighborhood size of k-nearest-neighbor. By modifying our implemented variants of the Sunny and 3S approaches in this way, we achieve strong performance on many ASlib benchmarks and establish new state-of-the-art performance on 3 scenarios.

Michael Brenner

• Marc Hanheide, Nick Hawes, Jeremy Wyatt, Moritz Göbelbecker, Michael Brenner, Kristoffer Sjöö, Alper Aydemir, Patric Jensfelt, Hendrik Zender and Geert-Jan Kruijff.
  A Framework for Goal Generation and Management.
  In Proceedings of the AAAI Workshop on Goal-Directed Autonomy. 2010.
  (Show abstract) (Hide abstract) (BIB)
  

  Goal-directed behaviour is often viewed as an essential char- acteristic of an intelligent system, but mechanisms to generate and manage goals are often overlooked. This paper addresses this by presenting a framework for autonomous goal gener- ation and selection. The framework has been implemented as part of an intelligent mobile robot capable of exploring unknown space and determining the category of rooms au- tonomously. We demonstrate the efficacy of our approach by comparing the performance of two versions of our inte- grated system: one with the framework, the other without. This investigation leads us conclude that such a framework is desirable for an integrated intelligent system because it re- duces the complexity of the problems that must be solved by other behaviour-generation mechanisms, it makes goal- directed behaviour more robust in the face of a dynamic and unpredictable environments, and it provides an entry point for domain-specific knowledge in a more general system.

• Michael Brenner.
  Creating Dynamic Story Plots with Continual Multiagent Planning.
  In Maria Fox and David Poole (eds.), Proceedings of the Twenty-Fourth AAAI Conference on Artificial Intelligence (AAAI 2010), pp. 1517-1522. AAAI Press 2010.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  An AI system that is to create a story (autonomously or in interaction with human users) requires capabilities from many subfields of AI in order to create characters that themselves appear to act intelligently and believably in a coherent story world. Specifically, the system must be able to reason about the physical actions and verbal interactions of the characters as well as their perceptions of the world. Furthermore it must make the characters act believably--i.e. in a goal-directed yet emotionally plausible fashion. Finally, it must cope with (and embrace!) the dynamics of a multiagent environment where beliefs, sentiments, and goals may change during the course of a story and where plans are thwarted, adapted and dropped all the time. In this paper, we describe a representational and algorithmic framework for modelling such dynamic story worlds, Continual Multiagent Planning. It combines continual planning (i.e. an integrated approach to planning and execution) with a rich description language for modelling epistemic and affective states, desires and intentions, sensing and communication. Analysing story examples generated by our implemented system we show the benefits of such an integrated approach for dynamic plot generation.

• Michael Brenner.
  Dynamic Plot Generation by Continual Multiagent Planning (extended abstract).
  In Proceedings of the 9th Int. Joint Conf. on Autonomous Agents and Multiagent Systems (AAMAS 2010). 2010.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  We describe how, by modelling plot generation as a Continual Multiagent Planning process, dynamic stories can be generated in which characters not only inteleave perception, action and interaction, but in which also beliefs and motivations may change repeatedly, thus driving the plot forward.

• Moritz Göbelbecker, Thomas Keller, Patrick Eyerich, Michael Brenner and Bernhard Nebel.
  Coming Up with Good Excuses: What To Do When No Plan Can be Found.
  In Ronen Brafman, Héctor Geffner, Jörg Hoffmann and Henry Kautz (eds.), Proceedings of the 20th International Conference on Automated Planning and Scheduling (ICAPS 2010), pp. 81-88. AAAI Press 2010.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  When using a planner-based agent architecture, many things can go wrong. First and foremost, an agent might fail to execute one of the planned actions for some reasons. Even more annoying, however, is a situation where the agent is incompetent, i.e., unable to come up with a plan. This might be due to the fact that there are principal reasons that prohibit a successful plan or simply because the task's description is incomplete or incorrect. In either case, an explanation for such a failure would be very helpful. We will address this problem and provide a formalization of coming up with excuses for not being able to find a plan. Based on that, we will present an algorithm that is able to find excuses and demonstrate that such excuses can be found in practical settings in reasonable time.

• Nick Hawes, Marc Hanheide, Kristoffer Sjöö, Alper Aydemir, Patric Jensfelt, Moritz Göbelbecker, Michael Brenner, Hendrik Zender, Pierre Lison, Ivana Kruijff-Korbayov, Geert-Jan M. Kruijff and Michael Zillich.
  Dora The Explorer: A Motivated Robot.
  In Proc. of 9th Int. Conf. on Autonomous Agents and Multiagent Systems (AAMAS 2010). 2010.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  Dora the Explorer is a mobile robot with a sense of curios- ity and a drive to explore its world. Given an incomplete tour of an indoor environment, Dora is driven by internal motivations to probe the gaps in her spatial knowledge. She actively explores regions of space which she hasn't previously visited but which she expects will lead her to further unex- plored space. She will also attempt to determine the cate- gories of rooms through active visual search for functionally important objects, and through ontology-driven inference on the results of this search.

• Christian Dornhege, Patrick Eyerich, Thomas Keller, Sebastian Trüg, Michael Brenner and Bernhard Nebel.
  Semantic Attachments for Domain-Independent Planning Systems.
  In Proceedings of the 19th International Conference on Automated Planning and Scheduling (ICAPS 2009), pp. 114-121. AAAI Press 2009.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  Solving real-world problems using symbolic planning often requires a simplified formulation of the original problem, since certain subproblems cannot be represented at all or only in a way leading to inefficiency. For example, manipulation planning may appear as a subproblem in a robotic planning context or a packing problem can be part of a logistics task. In this paper we propose an extension of PDDL for specifying semantic attachments. This allows the evaluation of grounded predicates as well as the change of fluents by externally specified functions. Furthermore, we describe a general schema of integrating semantic attachments into a forward-chaining planner and report on our experience of adding this extension to the planners FF and Temporal Fast Downward. Finally, we present some preliminary experiments using semantic attachments.

• Michael Brenner and Bernhard Nebel.
  Continual Planning and Acting in Dynamic Multiagent Environments.
  Journal of Autonomous Agents and Multiagent Systems 19 (3), pp. 297-331. 2009.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  In order to behave intelligently, artificial agents must be able to deliberatively plan their future actions. Unfortunately, realistic agent environments are usually highly dynamic and only partially observable, which makes planning computationally hard. For most practical purposes this rules out planning techniques that account for all possible contingencies in the planning process. However, many agent environments permit an alternative approach, namely continual planning, i.e. the interleaving of planning with acting and sensing.

  This paper presents a new principled approach to continual planning that describes why and when an agent should switch between planning and acting. The resulting continual planning algorithm enables agents to deliberately postpone parts of their planning process and instead actively gather missing information that is relevant for the later refinement of the plan. To this end, the algorithm explictly reasons about the knowledge (or lack thereof) of an agent and its sensory capabilities. These concepts are modelled in the planning language MAPL. Since in many environments the major reason for dynamism is the behaviour of other agents, MAPL can also model multiagent environments, common knowledge among agents, and communicative actions between them. For Continual Planning, MAPL introduces the concept of of assertions, abstract actions that substitute yet unformed subplans.

  To evaluate our continual planning approach empirically we have developed MAPSIM, a simulation environment that automatically builds multiagent simulations from formal MAPL domains. Thus, agents can not only plan, but also execute their plans, perceive their environment, and interact with each other. Our experiments show that, using continual planning techniques, deliberate action planning can be used efficiently even in complex multiagent environments.

• Geert-Jan Kruijff and Michael Brenner.
  Phrasing Questions.
  In AAAI Spring Symposium on Agents that Learn from Human Teachers. 2009.
  

• Michael Brenner.
  Continual Collaborative Planning for Mixed-Initiative Action and Interaction.
  In Proceedings of the 7th International Conference on Autonomous Agents and Multiagent Systems (AAMAS 2008). 2008.
  (Show abstract) (Hide abstract) (PDF)
  

  Multiagent environments are often highly dynamic and only partially observable which makes deliberative action planning computationally hard. In many such environments, however, agents can take a more proactive approach and suspend planning for partial plan execution, especially for active information gathering and interaction with others. This paper presents a new algorithm for Continual Collaborative Planning (CCP) that enables agents to deliberately interleave planning, acting, perception and communication. Our implementation of CCP has been evaluated with MAPSIM, a tool that automatically generates multiagent simulations from formal multiagent planning (MAP) domains. For different such simulations, we show how CCP leads to collaborative planning and acting and, despite minimal linguistic capabilities, to fairly natural dialogues between agents.

• Michael Brenner and Ivana Kruijff-Korbayova.
  A Continual Multiagent Planning Approach to Situated Dialogue.
  In Proceedings of the 12th Workshop on the Semantics and Pragmatics of Dialogue (LonDial 2008). 2008.
  (Show abstract) (Hide abstract) (PDF)
  

  Situated dialogue is usually tightly integrated with behavior planning, physical action and perception. This paper presents an algorithmic framework, Continual Collaborative Planning (CCP), for modeling this kind of integrated behavior and shows how CCP agents naturally blend physical and communicative actions. For experiments with conversational CCP agents we have developed MAPSIM, a software environment that can generate multiagent simulations from formal multiagent planning problems automatically. MAPSIM permits comparison of CCP-based dialogue strategies on a wide range of domains and problems without domain-specific programming. Despite their linguistic capabilities being limited MAPSIM agents can already engage in fairly realistic situated dialogues. Our ongoing work is taking this approach from simulation to real human-robot interaction.

• Geert-Jan Kruijff, Michael Brenner and Nick Hawes.
  Continual Planning for Cross-Modal Situated Clarification in Human-Robot Interaction.
  In Proceedings of the 17th IEEE International Symposium on Robots and Human Interactive Communication (RO-MAN 2008). 2008.
  (Show abstract) (Hide abstract) (PDF)
  

  Current robots do not fully understand the world they are situated in, including what humans talk to them about. A fundamental problem in robotics is thus how a robot can clarify such a lack of understanding. This paper addresses the question of how a robot can create a plan for resolving a need for clarification. This paper characterizes situated clarification as an information need which may arise in any sensory-motoric modality interpreting the situated context of the robot, or any deliberative modality referring to that context. The paper then focuses on how, once a clarification need has been identified, the robot can create a plan in which one or more modalities are involved in resolving it. Modalities are involved on the basis of the types of information they can provide. These information types are identified in the ontologies the modalities use to interconnect their content with content of other modalities ("information fusion"). We take a continual approach to planning and execution monitoring. This provides the abiltity to re-plan depending on modality availability and success in resolving (part of) a clarification need. We illustrate our implementation of this approach with several examples from our system.

• Paul Plöger, Kai Pervölz, Christoph Mies, Patrick Eyerich, Michael Brenner and Bernhard Nebel.
  The DESIRE Service Robotics Initiative.
  Künstliche Intelligenz 08 (4), pp. 29-32. 2008.
  (Show abstract) (Hide abstract) (PDF)
  

  We present some advanced hardware units and an appropriate component based SW architecture for DESIRE. As an example we describe the integration of a enhanced AI task planner which allows for higher flexibility and dependability during complex task execution.

• Patrick Eyerich, Michael Brenner and Bernhard Nebel.
  On the Complexity of Planning Operator Subsumption.
  In Proceedings of the Eleventh International Conference on Principles of Knowledge Representation and Reasoning (KR 2008), pp. 518-527. AAAI Press 2008.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  Formal action models play a central role in several subfields of AI because they are used to model application domains, e.g., in automated planning. However, there are hitherto no automated methods for relating such domain models to each other, in particular for checking whether one is a specialization or generalization of the other. In this paper, we introduce two kinds of subsumption relations between operators, both of which are suitable for modeling and verifying hierarchies between actions and operators: applicability subsumption considers an action to be more general than another if the latter can be replaced by the first at each point in each sound sequence of actions; abstraction subsumption exploits relations between actions from an ontological point of view. For both kinds of subsumption, we prove complexity results for verifying operator subsumption in three important subclasses: The problems are NP-complete when the expressiveness of the operators is restricted to the well-known basic STRIPS formalism, Sigma_p_2-complete when we admit boolean logical operators and undecidable when the full power of the planning language ADL is permitted.

• Michael Brenner.
  Situation-Aware Interpretation, Planning and Execution of User Commands by Autonomous Robots.
  In Proceedings of the 16th IEEE International Symposium on Robots and Human Interactive Communication (ROMAN 2007). Jeju, Korea 2007.
  (PDF)
  

• Nick Hawes, Aaron Sloman, Jeremy Wyatt, Michael Zillich, Henrik Jacobsson, Geert-Jan Kruijff, Michael Brenner, Gregor Berginc and Danijel Skocaj.
  Towards an Integrated Robot with Multiple Cognitive Functions.
  In Proceedings of the 22nd Conference on Artificial Intelligence (AAAI 2007). Vancouver, Canada 2007.
  (PDF)
  

• Geert-Jan Kruijff and Michael Brenner.
  Modelling Spatio-Temporal Comprehension in Situated Human-Robot Dialogue as Reasoning About Intentions and Plans.
  In AAAI Spring Symposium on Intentions. 2007.
  

• Michael Brenner, Nick Hawes, John Kelleher and Jeremy Wyatt.
  Mediating Between Qualitative and Quantitative Representations for Task-Orientated Human-Robot Interaction.
  In Proceedings of the Twentieth International Joint Conference on Artificial Intelligence (IJCAI 2007). Hyderabad, India 2007.
  (PDF)
  

• Michael Brenner and Bernhard Nebel.
  Continual Planning and Acting in Dynamic Multiagent Environments.
  In Proceedings of the International Symposium on Practical Cognitive Agents and Robots. Perth, Australia 2006.
  (Show abstract) (Hide abstract) (PDF)
  

  In highly dynamic environments, e.g. multiagent systems, finding optimal action plans is practically impossible since individual agents lack important knowledge at planning time or this knowledge has become obsolete when a plan is executed. It is often more practical in such environments to enable agents to ac- tively extend their knowledge as part of their plans and then revise their decisions in light of these update. In this paper, we describe a new principled approach to Continual Planning, i.e. the integration of Planning, Execution and Monitoring. The algorithm deliberately postpones parts of the planning process to later stages in an agent’s plan-act-monitor cycle and automatically determines when to switch back to refining or revising a partly executed plan. To evaluate our (and others’) Continual Planning techniques we have developed a simulation environment where formal MA Planning domains are not only used by planning agents but also as the basis of the simulation model such that agents can not only plan, but execute actions and perceive their environment. Our experiments show that, using continual planning techniques, deliberate action planning can be used efficiently even in complex multiagent environments.

• Alexander Kleiner, Michael Brenner, Tobias Braeuer, Christian Dornhege, Moritz Göbelbecker, Matthias Luber, Johann Prediger, Joerg Stueckler and Bernhard Nebel.
  Successful Search and Rescue in Simulated Disaster Areas.
  In Proceedings of the International RoboCup Symposium '05. Osaka, Japan 2005.
  (Show abstract) (Hide abstract) (PDF)
  

  RoboCupRescue Simulation is a large-scale multi-agent simulation of urban disasters where, in order to save lives and minimize damage, rescue teams must effectively cooperate despite sensing and communication limitations. This paper presents the comprehensive search and rescue approach of the ResQ Freiburg team, the winner in the RoboCupRescue Simulation league at RoboCup 2004. Specific contributions include the predictions of travel costs and civilian lifetime, the efficient coordination of an active disaster space exploration, as well as an any-time rescue sequence optimization based on a genetic algorithm. We compare the performances of our team and others in terms of their capability of extinguishing fires, freeing roads from debris, disaster space exploration, and civilian rescue. The evaluation is carried out with information extracted from simulation log files gathered during RoboCup 2004. Our results clearly explain the success of our team, and also confirm the scientific approaches proposed in this paper.

• Michael Brenner, Nanda Wijermans, Timo Nuessle and Bart de Boer.
  Simulating and Controlling Civilian Crowds in Robocup Rescue.
  In RoboCup. Osaka, Japan 2005.
  Winner of the RoboCupRescue Infrastructure Competition 2005.
  

• Michael Brenner.
  Planning for Multiagent Environments: From Individual Perceptions to Coordinated Execution.
  In Workshop on Multiagent Planning and Scheduling (ICAPS 2005). Monterey, USA 2005.
  (PDF)
  

• Timo Nuessle, Alexander Kleiner and Michael Brenner.
  Approaching Urban Disaster Reality: The ResQ Firesimulator.
  In Proceedings of the International RoboCup Symposium '04. Lisbon, Portugal 2004.
  (PDF)
  

• Alexander Kleiner, Michael Brenner, Tobias Braeuer, Christian Dornhege, Moritz Göbelbecker, Matthias Luber, Johann Prediger and Joerg Stueckler.
  ResQ Freiburg: Team Description and Evaluation, Team Description Paper, Rescue Simulation League.
  In CDROM Proceedings of the International RoboCup Symposium '04. Lisbon, Portugal 2004.
  (PDF)
  

• Michael Brenner.
  Multiagent Planning with Partially Ordered Temporal Plans.
  In Proceedings of IJCAI'03. Acapulco, Mexico 2003.
  (PDF)
  

• Michael Brenner.
  A Multiagent Planning Language.
  In Workshop on PDDL (ICAPS 2003). Trento, Italy 2003.
  (PDF)
  

• Michael Brenner.
  A Formal Model for Planning with Time and Resources in Concurrent Domains.
  In Workshop on Planning with Resources (IJCAI 2001). Seattle, Washington, USA 2001.
  (PS.GZ)
  

Markus Büttner

• Markus Büttner and Jussi Rintanen.
  Satisfiability Planning with Constraints on the Number of Operators.
  In Proceedings of the Thirteenth International Conference of Automated Planning and Scheduling (ICAPS 2005). Monterey, Califonia, USA 2005.
  

• Markus Büttner.
  Enhanced Prefetching- and Caching Strategies for Single and Multi-Disk Systems.
  ACTA INFORMATICA 236. 2005.
  

Patrick Caspari

• Patrick Caspari, Robert Mattmüller and Tim Schulte.
  A Framework to Prove Strong Privacy in Multi-Agent Planning.
  In Proceedings of the 6th Workshop on Distributed and Multi-Agent Planning (DMAP 2020), pp. 32-39. 2020.
  (PDF)
  

Yannis Dimopoulos

• Yannis Dimopoulos, Bernhard Nebel and Francesca Toni.
  On the Computational Complexity of Assumption-based Argumentation for Default Reasoning.
  Artificial Intelligence 141 (1-2), pp. 57-78. 2002.
  (Show abstract) (Hide abstract) (PDF)
  

  Bondarenko et al. have recently proposed an abstract framework for default reasoning. Besides capturing most existing formalisms and proving that their standard semantics all coincide, the framework extends these formalisms by generalising the semantics of admissible and preferred arguments, originally proposed for logic programming only. In this paper we analyse the computational complexity of credulous and sceptical reasoning under the semantics of admissible and preferred arguments for (the propositional variant of) the instances of the abstract framework capturing theorist, circumscription, logic programming, default logic, and autoepistemic logic. Although the new semantics have been tacitly assumed to mitigate the computational hardness of default reasoning under the standard semantics of stable extensions, we show that in many cases reasoning under the admissibility and preferability semantics is computationally harder than under the standard semantics. In particular, in the case of autoepistemic logic, sceptical reasoning under preferred arguments is located at the fourth level of the polynomial hierarchy, whereas the same form of reasoning under stable extensions is located at the second level.

• Yannis Dimopoulos, Bernhard Nebel and Francesca Toni.
  Finding Admissible and Preferred Arguments Can be Very Hard.
  In Principles of Knowledge Representation and Reasoning, Proceedings of the 7th International Conference (KR'2000). 2000.
  (Show abstract) (Hide abstract) (PDF)
  

  Bondarenko et al. have recently proposed an extension of the argumentation-theoretic semantics of admissible and preferred arguments, originally proposed for logic programming only, to a number of other nonmonotonic reasoning formalisms. In this paper we analyse the computational complexity of credulous and sceptical reasoning under the semantics of admissible and preferred arguments for (the propositional variant of) some well-known frameworks for nonmonotonic reasoning, i.e. Theorist, Circumscription and Autoepistemic Logic. While the new semantics have been assumed to mitigate the computational problems of nonmonotonic reasoning under the standard semantics of stable extensions, we show that in many cases reasoning under the new semantics is computationally harder than under the standard semantics. In particular, for Autoepistemic Logic, the sceptical reasoning problem under the semantics of preferred arguments is located at the fourth level of the polynomial hierarchy, two levels above the same problem under the standard semantics. In some cases, however, reasoning under the new semantics becomes easier - reducing to reasoning in the monotonic logics underlying the nonmonotonic frameworks.

• Yannis Dimopoulos, Bernhard Nebel and Francesca Toni.
  Preferred Arguments are Harder to Compute than Stable Extensions.
  In Proceedings of the 16th International Joint Conference on Artificial Intelligence (IJCAI 1999). Stockholm, Sweden 1999.
  (Show abstract) (Hide abstract) (PDF)
  

  Based on an abstract framework for nonmonotonic reasoning, Bondarenko et al. have extended the logic programming semantics of admissible and preferred arguments to other nonmonotonic formalisms such as circumscription, auto-epistemic logic and default logic. Although the new semantics have been tacitly assumed to mitigate the computational problems of nonmonotonic reasoning under the standard semantics of stable extensions, it seems questionable whether they improve the worst-case behaviour. As a matter of fact, we show that credulous reasoning under the new semantics in propositional logic programming and propositional default logic has the same computational complexity as under the standard semantics. Furthermore, sceptical reasoning under the admissibility semantics is easier - since it is trivialised to monotonic reasoning. Finally, sceptical reasoning under the preferability semantics is harder than under the standard semantics.

• Yannis Dimopoulos, Bernhard Nebel and Jana Koehler.
  Encoding planning problems in non-monotonic logic programs.
  In Proc. European Conference on Planning 1997 (ECP-97), pp. 169-181. Springer-Verlag 1997.
  (Show abstract) (Hide abstract) (PDF)
  

  In this paper we study a framework for encoding planning problems in logic programs with negation as failure. In contrast to other research work that focuses on the representional adequacy of nonmontonic logic programming as a language for describing theories of action and change, here we are concerned with more practical issues. Namely, we examine the effectiveness of an existing implementation of the stable models semantics called "Smodels" in solving a series of hard planning problems. We discuss representational issues and point out factors that can influence the performance of the method. It turns out that for careful and compact encodings, the performance of the method in a number of different domains, is comparable to that of planners like GRAPHPLAN and SATPLAN.

• Yannis Dimopoulos, Saso Dzeroski and Antonis Kakas.
  Integrating Explanatory and Descriptive Learning in ILP.
  In Proceedings of the 15th International Joint Conference on Artificial Intelligence (IJCAI'97), pp. 900-906. 1997.
  (PS.GZ)
  

• Jana Koehler, Bernhard Nebel, Jörg Hoffmann and Yannis Dimopoulos.
  Extending Planning Graphs to an ADL Subset.
  In Proc. European Conference on Planning 1997 (ECP-97), pp. 273-285. Springer-Verlag 1997.
  (Show abstract) (Hide abstract) (PDF)
  

  We describe an extension of GRAPHPLAN to a subset of ADL that allows conditional and universally quantified effects in operators in such a way that almost all interesting properties of the original Graphplan algorithm are preserved.

• Bernhard Nebel, Yannis Dimopoulos and Jana Koehler.
  Ignoring Irrelevant Facts and Operators in Plan Generation.
  In Proc. European Conference on Planning 1997 (ECP-97), pp. 338-350. Springer-Verlag 1997.
  (Show abstract) (Hide abstract) (PDF)
  

  It is traditional wisdom that one should start from the goals when generating a plan in order to focus the plan generation process on potentially relevant actions. The GRAPHPLAN system, however, which is the most efficient planning system nowadays, builds a ``planning graph'' in a forward-chaining manner. Although this strategy seems to work well, it may possibly lead to problems if the planning task description contains irrelevant information. Although some irrelevant information can be filtered out by GRAPHPLAN, most cases of irrelevance are not noticed.

  In this paper, we analyze the effects arising from ``irrelevant'' information to planning task descriptions for different types of planners. Based on that, we propose a family of heuristics that select relevant information by minimizing the number of initial facts that are used when approximating a plan by backchaining from the goals ignoring any conflicts. These heuristics, although not solution-preserving, turn out to be very useful for guiding the planning process, as shown by applying the heuristics to a large number of examples from the literature.

Christian Dornhege

• David Speck, Christian Dornhege and Wolfram Burgard.
  Shakey 2016 - How Much Does it Take to Redo Shakey the Robot?
  IEEE Robotics and Automation Letters (RA-L) 2 (2), pp. 1203-1209. 2017.
  (Show abstract) (Hide abstract) (PDF; Online)
  

  Shakey the robot was one of the first autonomous robots that showed impressive capabilities of navigation and mobile manipulation. Since then, robotics research has made great progress, showing more and more capable robotic systems for a large variety of application domains and tasks. In this letter, we look back on decades of research by rebuilding Shakey with modern robotics technology in the open-source Shakey 2016 system. Hereby, we demonstrate the impact of research by showing that ideas from the original Shakey are still alive in state-of-the-art systems, while robotics in general has improved to deliver more robust and more capable software and hardware. Our Shakey 2016 system has been implemented on real robots and leverages mostly open-source software. We experimentally evaluate the system in real-world scenarios on a PR2 robot and a Turtlebot-based robot and particularly investigate the development effort. The experiments documented in this letter demonstrate that results from robotics research are readily available for building complex robots such as Shakey within a short amount of time and little effort.

• Christian Dornhege, Alexander Kleiner, Andreas Hertle and Andreas Kolling.
  Multirobot Coverage Search in Three Dimensions.
  Journal of Field Robotics 33 (4), pp. 537-558. 2016.
  (Show abstract) (Hide abstract) (PDF)
  

  Searching for objects and observing parts of a known environment efficiently is a fundamental problem in many real-world robotic applications, e.g., household robots searching for objects, inspection robots searching for leaking pipelines, and rescue robots searching for survivors after a disaster. We consider the problem of identifying and planning sequences of sensor locations from which robot sensors can observe and cover complex three-dimensional (3D) environments while traveling only short distances. Our approach is based on sampling and ranking a large number of sensor locations for a 3D environment represented by an OctoMap. The visible area from these sensor locations induces a minimal partition of the 3D environment that we exploit for planning sequences of sensor locations with short travel times efficiently. We present multiple planning algorithms designed for single robots and for multirobot teams. These algorithms include variants that are greedy, optimal, or based on decomposing the planning problem into a set cover and traveling salesman problem. We evaluated and compared these algorithms empirically in simulation and real-world robot experiments with up to four robots. Our results demonstrate that, despite the intractability of the overall problem, computing and executing effective solutions for multirobot coverage search in real 3D environments is feasible and ready for real-world applications.

• Armin Hornung, Sebastian Boettcher, Christian Dornhege, Andreas Hertle, Jonas Schlagenhauf and Maren Bennewitz.
  Mobile Manipulation in Cluttered Environments with Humanoids: Itegrated Perception, Task Planning, and Action Execution.
  In Proceedings of the IEEE-RAS International Conference on Humanoid Robots (HUMANOIDS). 2014.
  (Show abstract) (Hide abstract) (PDF)
  

  To autonomously carry out complex mobile manipulation tasks, a robot control system has to integrate several components for perception, world modeling, action planning and replanning, navigation, and manipulation. In this paper, we present a modular framework that is based on the Temporal Fast Downward Planner and supports external modules to control the robot. This allows to tightly integrate individual sub-systems with the high-level symbolic planner and enables a humanoid robot to solve challenging mobile manipulation tasks. In the work presented here, we address mobile manipulation with humanoids in cluttered environments, particularly the task of collecting objects and delivering them to designated places in a home-like environment while clearing obstacles out of the way. We implemented our system for a Nao humanoid tidying up a room, i.e., the robot has to collect items scattered on the floor, move obstacles out of its way, and deliver the objects to designated target locations. Despite the limited sensing and motion capabilities of the low-cost platform, the experiments show that our approach results in reliable task execution by applying monitoring actions to verify object and robot states.

• Nicolas Riesterer, Christian Becker-Asano, Julien Hué, Christian Dornhege and Bernhard Nebel.
  The Hybrid Agent MARCO.
  In Proceedings of the 16th International Conference on Multimodal Interaction, pp. 80-81. 2014.
  (Show abstract) (Hide abstract) (PDF)
  

  We present MARCO, a hybrid, chess playing agent equipped with a custom-built robotic arm and a virtual agent’s face displaying emotions. MARCO was built to investigate the hypothesis that hybrid agents capable of displaying emotions make playing chess more personal and enjoyable. In addition, we aim to explore means of achieving emotional contagion between man and machine.

• Christian Becker-Asano, Eduardo Meneses, Nicolas Riesterer, Julien Hué, Christian Dornhege and Bernhard Nebel.
  The Hybrid Agent MARCO: A Multimodal Autonomous Robotic Chess Opponent.
  In Proceedings of the 2nd International Confernce on Human-Agent Interaction, pp. 173-176. 2014.
  (Show abstract) (Hide abstract) (PDF)
  

  We present MARCO, a hybrid, chess playing agent equipped with a custom-built robotic arm and a virtual agent’s face displaying emotions. MARCO was built to investigate the hypothesis that hybrid agents capable of displaying emo- tions make playing chess more personal and enjoyable. In addition, we aim to explore means of achieving emotional contagion between man and machine.

• Andreas Hertle, Christian Dornhege, Thomas Keller, Robert Mattmüller, Manuela Ortlieb and Bernhard Nebel.
  An Experimental Comparison of Classical, FOND and Probabilistic Planning.
  In Proceedings of the 37th German Conference on Artificial Intelligence (KI 2014), pp. 297-308. Springer 2014.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  Domain-independent planning in general is broadly applicable to a wide range of tasks. Many formalisms exist that allow to describe different aspects of realistic problems. Which one is best is not clear as usually more expressiveness comes with a cost in planning time. Under the assumption that hard guarantees are not required, users are faced with a decision between multiple approaches. In this work we study the effect of nondeterministic action outcomes. As a generic model we use a probabilistic description in the form of Markov Decision Processes (MDPs). We define abstracting translations into a classical planning formalism and fully observable nondeterministic (FOND) planning. Our goal is to give insight into how state-of-the-art systems perform on different MDP planning domains. We evaluate those MDPs by running state-of-the-art planning systems on the abstracted formalisms.

• Christian Dornhege, Andreas Hertle and Bernhard Nebel.
  Lazy Evaluation and Subsumption Caching for Search-Based Integrated Task and Motion Planning.
  In Proceedings of the IROS workshop on AI-based robotics. 2013.
  (Show abstract) (Hide abstract) (PDF)
  

  State of the art classical planning systems can efficiently solve large symbolic problem instances. Applying classical planning techniques to robotics is possible by in- tegrating geometric reasoning in the planning process. The problems that are solvable in this way are significantly smaller than purely logical formulations as many costly geometric calculations are requested by a planner. Therefore we aim to avoid those calculations while preserving correctness. We address this problem with efficient caching techniques. Subsumption caching avoids costly computations by caching geometric queries and beyond answering the same queries also considers less or more constrained ones. Additionally, we describe a lazy evaluation technique that pushes applicability checks for successor states performing geometric queries to a later point. As we are interested in the performance of our planner not as a standalone component, but as part of an intelligent robotic system, we evaluate those techniques embedded in an integrated system during real-world mobile manipulation experiments.

• Christian Dornhege, Alexander Kleiner and Andreas Kolling.
  Coverage Search in 3D.
  In Proceedings of the Symposium on Safety Security and Rescue Robotics (SSRR). 2013.
  (PDF) (BIB)
  

• Andreas Hertle and Christian Dornhege.
  Efficient Extensible Path Planning on 3D Terrain Using Behavior Modules.
  In Proceedings of the European Conference on Mobile Robotics (ECMR). 2013.
  (PDF) (BIB)
  

• Bernhard Nebel, Christian Dornhege and Andreas Hertle.
  How Much Does a Household Robot Need To Know In Order To Tidy Up Your Home?
  In AAAI Workshop on Intelligent Robotic Systems. AAAI Press 2013.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  Although planning for the tasks a household robot has to perform appears to be easy, there exists the problem that the robot is usually uncertain about the state of the household when starting to plan. For example, when getting the order of tidying up the kitchen, the robot does not know what objects it will have to put away and whether there are actually any objects that need to be put away. Furthermore, while sensing operations can provide more information about the environment, things can go wrong when executing an action. In this paper, we try to identify conditions under which classical planning can be used in a replanning loop in order to solve the planning problem in nondeterminis- tic partially observable open domains. In particular, we will define completeness and soundness of replanning with respect to nondeterministic planning and we will identify a PSPACE-checkable condition that guarantees soundness.

• Christian Dornhege and Andreas Hertle.
  Integrated Symbolic Planning in the Tidyup-Robot Project.
  In AAAI Spring Symposium - Designing Intelligent Robots: Reintegrating AI II. AAAI Press 2013.
  (PDF) (BIB)
  

• Christian Dornhege and Alexander Kleiner.
  A Frontier-Void-Based Approach for Autonomous Exploration in 3D.
  Advanced Robotics 27 (6). 2013.
  (BIB)
  

• Kai M. Wurm, Christian Dornhege, Cyrill Stachniss, Bernhard Nebel and Wolfram Burgard.
  Coordinating Heterogeneous Teams of Robots using Temporal Symbolic Planning.
  Autonomous Robots. 2013.
  (Show abstract) (Hide abstract) (BIB) (Online; DOI)
  

  The efficient coordination of a team of heterogeneous robots is an important requirement for exploration, rescue, and disaster recovery missions. In this paper, we present a novel approach to target assignment for heterogeneous teams of robots. It goes beyond existing target assignment algorithms in that it explicitly takes symbolic actions into account. Such actions include the deployment and retrieval of other robots or manipulation tasks. Our method integrates a temporal planning approach with a traditional cost-based planner. The proposed approach was implemented and evaluated in two distinct settings. First, we coordinated teams of marsupial robots. Such robots are able to deploy and pickup smaller robots. Second, we simulated a disaster scenario where the task is to clear blockades and reach certain critical locations in the environment. A similar setting was also investigated using a team of real robots. The results show that our approach outperforms ad-hoc extensions of state-of-the-art cost-based coordination methods and that the approach is able to efficiently coordinate teams of heterogeneous robots and to consider symbolic actions.

• Stefan Kohlbrecher, Karen Petersen, Gerald Steinbauer, Johannes Maurer, Peter Lepej, Suzana Uran, Rodrigo Ventura, Christian Dornhege, Andreas Hertle, Raymond Sheh and Johannes Pellenz.
  Community-Driven Development of Standard Software Modules for Search and Rescue Robots.
  In Safety, Security and Rescue Robotics (SSRR). 2012.
  

• Andreas Hertle, Christian Dornhege, Thomas Keller and Bernhard Nebel.
  Planning with Semantic Attachments: An Object-Oriented View.
  In Proceedings of the European Conference on Artificial Intelligence (ECAI). 2012.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  In recent years, domain-independent planning has been applied to a rising number of real-world applications. Usually, the description language of choice is PDDL. However, PDDL is not suited to model all challenges imposed by real-world applications. Dornhege et al. proposed semantic attachments to allow the computation of Boolean fluents by external processes called modules during planning. To acquire state information from the planning system a module developer must perform manual requests through a callback interface which is both inefficient and error-prone. In this paper, we present the Object-oriented Planning Language OPL, which incorporates the structure and advantages of modern object-oriented programming languages. We demonstrate how a domain-specific module interface that allows to directly access the planner state using object member functions is automatically gen- erated from an OPL planning task. The generated domain-specific interface allows for a safe and less error-prone implementation of modules. We show experimentally that this interface is more efficient than the PDDL-based module interface of TFD/M.

• Christian Dornhege and Alexander Kleiner.
  A Frontier-Void-Based Approach for Autonomous Exploration in 3D.
  In Proceedings of the IEEE International Symposium on Safety, Security and Rescue Robotics (SSRR). 2011.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  We consider the problem of an autonomous robot searching for objects in unknown 3d space. Similar to the well known frontier-based exploration in 2d, the problem is to determine a minimal sequence of sensor viewpoints until the entire search space has been explored. We introduce a novel approach that combines the two concepts of voids, which are unexplored volumes in 3d, and frontiers, which are regions on the boundary between voids and explored space. Our approach has been evaluated on a mobile platform equipped with a manipulator searching for victims in a simulated USAR setup. First results indicate the real-world capability and search efficiency of the proposed method.

• Matthias Westphal, Christian Dornhege, Stefan Wölfl, Marc Gissler and Bernhard Nebel.
  Guiding the Generation of Manipulation Plans by Qualitative Spatial Reasoning.
  Spatial Cognition & Computation: An Interdisciplinary Journal 11 (1), pp. 75-102. 2011.
  (Show abstract) (Hide abstract) (Online; DOI) (BIB)
  

  Manipulation planning is a complex task for robots with a manipulator arm that need to grasp objects in the environment, specifically under narrow spatial conditions restricting the workspace of the robot. A popular approach for generating motion plans is probabilistic roadmap planning. However, the sampling strategy of such planners is usually unguided, and hence may lead to motion plans that seem counterintuitive for a human observer. In this article we present an approach that generates heuristics for the probabilistic sampling strategy from spatial plans that abstract from concrete metric data. These spatial plans describe a free trajectory in the workspace of the robot on a purely qualitative level, i.e., by employing spatial relations from formalisms considered in the domain of Qualitative Spatial and Tem- poral Reasoning. We discuss how such formalisms and constraint-based reasoning methods can be applied to approximate geometrically feasible motions. The paper is completed by an evaluation of a hybrid planning system in different spatial settings showing that run-times are notably improved when an abstract plan is considered as a guidance heuristic.

• R. Kümmerle, B. Steder, Christian Dornhege, Alexander Kleiner, G. Grisetti and W. Burgard.
  Large Scale Graph-based SLAM using Aerial Images as Prior Information.
  Autonomous Robots 30, pp. 25-39. 2011.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  The problem of learning a map with a mobile robot has been intensively studied in the past and is usually referred to as the simultaneous localization and mapping (SLAM) problem. However, most existing solutions to the SLAM problem learn the maps from scratch and have no means for incorporating prior information. In this paper, we present a novel SLAM approach that achieves global consistency by utilizing publicly accessible aerial photographs as prior information. It inserts correspondences found between stereo and three-dimensional range data and the aerial images as constraints into a graph-based formulation of the SLAM problem. We evaluate our algorithm based on large real-world datasets acquired even in mixed in- and outdoor environments by comparing the global accuracy with state-of-the-art SLAM approaches and GPS. The experimental results demonstrate that the maps acquired with our method show increased global consistency.

• Kai M. Wurm, Christian Dornhege, Patrick Eyerich, Cyrill Stachniss, Bernhard Nebel and Wolfram Burgard.
  Coordinated Exploration with Marsupial Teams of Robots using Temporal Symbolic Planning.
  In Proceedings of the 2010 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS 2010). 2010.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  The problem of autonomously exploring an environment with a team of robots received considerable attention in the past. However, there are relatively few approaches to coordinate teams of robots that are able to deploy and retrieve other robots. Efficiently coordinating the exploration with such marsupial robots requires advanced planning mechanisms that are able to consider symbolic deployment and retrieval actions. In this paper, we propose a novel approach for coordinating the exploration with marsupial robot teams. Our method integrates a temporal symbolic planner that explicitly considers deployment and retrieval actions with a traditional cost-based assignment procedure. Our approach has been implemented and evaluated in several simulated environments and with varying team sizes. The results demonstrate that our proposed method is able to coordinate marsupial teams of robots to efficiently explore unknown environments.

• Alexander Kleiner and Christian Dornhege.
  Mapping for the Support of First Responders in Critical Domains.
  Journal of Intelligent and Robotic Systems (JINT), pp. 1-29. 2010.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  In critical domains such as urban search and rescue (USAR), and bomb disposal, the deployment of teleoperated robots is essential to reduce the risk of first responder personnel. Teleoperation is a difficult task, particularly when controlling robots from an isolated safety zone. In general, the operator has to solve simultaneously the problems of mission planning, target identification, robot navigation, and robot control. We introduce a system to support teleoperated navigation with real-time mapping consisting of a two-step scan matching method that re-considers data associations during the search. The algorithm processes data from laser range finder and gyroscope only, thereby it is independent from the robot platform. Furthermore, we introduce a user-guided procedure for improving the global consistency of maps generated by the scan matcher. Globally consistent maps are computed by a graph-based maximum likelihood method that is biased by localizing crucial parts of the scan matcher trajectory on a prior given geo-tiff image. The approach has been implemented as an embedded system and extensively tested on robot platforms designed for teleoperation in critical situations, such as bomb disposal. Furthermore, the system was evaluated in a test maze by first responders during the Disaster City event in Texas 2008.

• Marc Gissler, Christian Dornhege, Bernhard Nebel and Matthias Teschner.
  Deformable Proximity Queries and their Application in Mobile Manipulation Planning.
  In Symposium on Visual Computing (ISVC 2009), pp. 79-88. AAAI Press 2009.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  We describe a proximity query algorithm for the exact minimum distance computation between arbitrarily shaped objects. Special characteristics of the Gilbert-Johnson-Keerthi (GJK) algorithm are employed in various stages of the algorithm. In the first stage, they are used to search for sub-mesh pairs whose convex hulls do not intersect. In the case of an intersection, they guide a recursive decomposition. Finally, they are used to derive lower and upper distance bounds in non-intersecting cases. These bounds are utilized in a spatial subdivision scheme to achieve a twofold culling of the domain. The algorithm does not depend on spatial or temporal coherence and is, thus, specifically suited to be applied to deformable objects. Furthermore, we describe its embedding into the geometrical part of a mobile manipulation planning system. Experiments show its usability in dynamic scenarios with deformable objects as well as in complex manipulation planning scenarios.

• Alexander Kleiner and Christian Dornhege.
  Operator-Assistive Mapping in Harsh Environments.
  In IEEE International Workshop on Safety, Security and Rescue Robotics (SSRR 2009), pp. 1-6. 2009.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  Teleoperation is a difficult task, particularly when controlling robots from an isolated operator station. In general, the operator has to solve nearly blindly the problems of mission planning, target identification, robot navigation, and robot control at the same time. The goal of the proposed system is to support teleoperated navigation with real-time mapping. We present a novel scan matching technique that re-considers data associations during the search, enabling robust pose estimation even under varying roll and pitch angle of the robot enabling mapping on rough terrain. The approach has been implemented as an embedded system and extensively tested on robot platforms designed for teleoperation in critical situations, such as bomb disposal. Furthermore, the system has been evaluated in a test maze by first responders during the Disaster City event in Texas 2008. Finally, experiments conducted within different environments show that the system yields comparably accurate maps in real-time when compared to higher sophisticated offline methods, such as Rao-Blackwellized SLAM.

• Christian Dornhege, Marc Gissler, Matthias Teschner and Bernhard Nebel.
  Integrating Symbolic and Geometric Planning for Mobile Manipulation.
  In IEEE International Workshop on Safety, Security and Rescue Robotics (SSRR 2009). 2009.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  Mobile manipulation requires to solve multiple subproblems. One is planning in high-dimensional configuration spaces, that we approach in this work. We decompose the manipulation problem into a symbolic and a geometric part. The symbolic part is implemented as a classical symbolic planner that tightly integrates a geometric planner enabling us to efficiently generate correct plans. A probabilistic roadmap planner constitutes the geometric part. During the computation of the roadmap we utilize proximity queries to determine non-colliding configurations and to verify collision-free paths between configurations accurately and efficiently. We demonstrate experiments in two scenarios, one of these being the manipulator dexterity test scenario that was used in NIST's response robot evaluation in Disaster City.

• Christian Dornhege, Patrick Eyerich, Thomas Keller, Sebastian Trüg, Michael Brenner and Bernhard Nebel.
  Semantic Attachments for Domain-Independent Planning Systems.
  In Proceedings of the 19th International Conference on Automated Planning and Scheduling (ICAPS 2009), pp. 114-121. AAAI Press 2009.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  Solving real-world problems using symbolic planning often requires a simplified formulation of the original problem, since certain subproblems cannot be represented at all or only in a way leading to inefficiency. For example, manipulation planning may appear as a subproblem in a robotic planning context or a packing problem can be part of a logistics task. In this paper we propose an extension of PDDL for specifying semantic attachments. This allows the evaluation of grounded predicates as well as the change of fluents by externally specified functions. Furthermore, we describe a general schema of integrating semantic attachments into a forward-chaining planner and report on our experience of adding this extension to the planners FF and Temporal Fast Downward. Finally, we present some preliminary experiments using semantic attachments.

• Moritz Göbelbecker and Christian Dornhege.
  Realistic Cities in Simulated Environments - An Open Street Map to Robocup Rescue Converter.
  In Online-Proceedings of the Fourth International Workshop on Synthetic Simulation and Robotics to Mitigate Earthquake Disaster (SRMED 2009). 2009.
  (Show abstract) (Hide abstract) (PDF)
  

  A general problem when developing large scale disaster simulation environments is to acquire GIS data. In this work, we tackle the problem of map generation from public sources. Usually the major problem is not only the data conversion itself, but to get access to the data at all. We solve this problem by using the website OpenStreetMap.org, that provides mapping data for the whole world in a wiki-style concept, as our source of data, thus being able to generate maps for almost any city. The data is converted to the format required by the Robocup Rescue Simulation System, enabling simulations on various real-world scenarios.

• Rainer Kümmerle, Bastian Steder, Christian Dornhege, Michael Ruhnke, Giorgio Grisetti, Cyrill Stachniss and Alexander Kleiner.
  On Measuring the Accuracy of SLAM Algorithms.
  Autonomous Robots 27 (4), pp. 387-407. 2009.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  In this paper, we address the problem of creating an objective benchmark for evaluating SLAM approaches. We propose a framework for analyzing the results of a SLAM approach based on a metric for measuring the error of the corrected trajectory. This metric uses only relative relations between poses and does not rely on a global reference frame. This overcomes serious shortcomings of approaches using a global reference frame to compute the error. Our method furthermore allows us to compare SLAM approaches that use different estimation techniques or different sensor modalities since all computations are made based on the corrected trajectory of the robot.

  We provide sets of relative relations needed to compute our metric for an extensive set of datasets frequently used in the robotics community. The relations have been obtained by manually matching laser-range observations to avoid the errors caused by matching algorithms. Our benchmark framework allows the user to easily analyze and objectively compare different SLAM approaches.

• Wolfram Burgard, Cyrill Stachniss, Giorgio Grisetti, Bastian Steder, Rainer Kümmerle, Christian Dornhege, Michael Ruhnke, Alexander Kleiner and Juan D. Tardos.
  A Comparison of SLAM Algorithms Based on a Graph of Relations.
  In Proceedings of the 2009 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS 2009), pp. 2089-2095. IEEE 2009.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  In this paper, we address the problem of creating an objective benchmark for comparing SLAM approaches. We propose a framework for analyzing the results of SLAM approaches based on a metric for measuring the error of the corrected trajectory. The metric uses only relative relations between poses and does not rely on a global reference frame. The idea is related to graph-based SLAM approaches in the sense that it considers the energy needed to deform the trajectory estimated by a SLAM approach to the ground truth trajectory. Our method enables us to compare SLAM approaches that use different estimation techniques or different sensor modalities since all computations are made based on the corrected trajectory of the robot. We provide sets of relative relations needed to compute our metric for an extensive set of datasets frequently used in the SLAM community. The relations have been obtained by manually matching laser-range observations. We believe that our benchmarking framework allows the user an easy analysis and objective comparisons between different SLAM approaches.

• Rainer Kümmerle, Bastian Steder, Christian Dornhege, Alexander Kleiner, Giorgio Grisetti and Wolfram Burgard.
  Large Scale Graph-based SLAM using Aerial Images as Prior Information.
  In Proceedings of 2009 Robotics: Science and Systems (RSS 2009). 2009.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  To effectively navigate in their environments and accurately reach their target locations, mobile robots require a globally consistent map of the environment. The problem of learning a map with a mobile robot has been intensively studied in the past and is usually referred to as the simultaneous localization and mapping (SLAM) problem. However, existing solutions to the SLAM problem typically rely on loop-closures to obtain global consistency and do not exploit prior information even if it is available. In this paper, we present a novel SLAM approach that achieves global consistency by utilizing publicly accessible aerial photographs as prior information. Our approach inserts correspondences found between three-dimensional laser range scans and the aerial image as constraints into a graph-based formulation of the SLAM problem. We evaluate our algorithm based on large real-world datasets acquired in a mixed in- and outdoor environment by comparing the global accuracy with state-of-the-art SLAM approaches and GPS. The experimental results demonstrate that the maps acquired with our method show increased global consistency.

• Christian Dornhege and Alexander Kleiner.
  Fully autonomous planning and obstacle negotiation on rough terrain using behavior maps.
  In Video Proceedings of the IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS 2007). San Diego, California 2007.
  

• Christian Dornhege and Alexander Kleiner.
  Behavior maps for online planning of obstacle negotiation and climbing on rough terrain.
  In Proceedings of the IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS 2007), pp. 3005-3011. 2007.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  To autonomously navigate on rough terrain is a challenging problem for mobile robots, requiring the ability to decide whether parts of the environment can be traversed or have to be bypassed, which is commonly known as Obstacle Negotiation (ON). In this paper, we introduce a planning framework that extends ON to the general case, where different types of terrain classes directly map to specific robot skills, such as climbing stairs and ramps. This extension is based on a new concept called behavior maps, which is utilized for the planning and execution of complex skills. Behavior maps are directly generated from elevation maps, i.e. two-dimensional grids storing in each cell the corresponding height of the terrain surface, and a set of skill descriptions. Results from extensive experiments are presented, showing that the method enables the robot to explore successfully rough terrain in real-time, while selecting the optimal trajectory in terms of costs for navigation and skill execution.

• Alexander Kleiner and Christian Dornhege.
  Real-time Localization and Elevation Mapping within Urban Search and Rescue Scenarios.
  Journal of Field Robotics 24, pp. 723-745. 2007.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  Urban Search And Rescue (USAR) is a time critical task. Rescue teams have to explore a large terrain within a short amount of time in order to locate survivors after a disaster. One goal in Rescue Robotics is to have a team of heterogeneous robots that explore autonomously, or partially guided by an incident commander, the disaster area. Their task is to jointly create a map of the terrain and to register victim locations, which can further be utilized by human task forces for rescue. Basically, the robots have to solve autonomously in real-time the problem of Simultaneous Localization and Mapping (SLAM), consisting of a continuous state estimation problem and a discrete data association problem. Extraordinary circumstances after a real disaster make it very hard to apply common techniques. Many of these have been developed under strong assumptions, for example, they require polygonal structures, such as typically found in office-like environments. Furthermore, most techniques are not deployable in real-time. In this paper we propose real-time solutions for localization and mapping, which all have been extensively evaluated within the test arenas of the National Institute of Standards and Technology (NIST). We specifically deal with the problems of vision-based pose tracking on tracked vehicles, the building of globally consistent maps based on a network of RFID tags, and the building of elevation maps from readings of a tilted Laser Range Finder (LRF). Our results show that these methods lead under modest computational requirements to good results within the utilized testing arenas.

• Alexander Kleiner, Christian Dornhege and Dali Sun.
  Mapping disaster areas jointly: RFID -Coordinated SLAM by Humans and Robots.
  In Proceedings of the IEEE International Workshop on Safety, Security and Rescue Robotics (SSRR 2007), pp. 1-6. 2007.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  We consider the problem of jointly performing SLAM by humans and robots in Urban Search And Rescue (USAR) scenarios. In this context, SLAM is a challenging task. First, places are hardly re-observable by vision techniques since visibility might be affected by smoke and fire. Second, loop-closure is cumbersome due to the fact that firemen will intentionally try to avoid performing loops when facing the reality of emergency response, e.g.USAR, while they are searching for victims. Furthermore, there might be places that are only accessible to robots, making it necessary to integrate humans and robots into one team for mapping the area after a disaster. In this paper, we introduce a method for jointly correcting individual trajectories of humans and robots by utilizing RFID technology for data association. Hereby the poses of humans and robots are tracked by a PDR (Pedestrian Dead Reckoning), and slippage sensitive odometry, respectively. We conducted extensive experiments with a team of humans, and a human-robot team within a semi-outdoor environment. Results from these experiments show that the introduced method allows to improve single trajectories based on the joint graph, even if they do not contain any loop.

• Alexander Kleiner, Christian Dornhege, Rainer Kuemmerle, Michael Ruhnke, Bastian Steder, Bernhard Nebel, Patrick Doherty, Mariusz Wzorek, Piotr Rudol, Gianpaolo Conte, S. Durante and D. Lundstrom.
  RoboCupRescue - Robot League Team RescueRobots Freiburg (Germany), Team Description Paper.
  In CDROM Proceedings of the International RoboCup Symposium '05. Bremen, Germany 2006.
  (Show abstract) (Hide abstract) (PDF)
  

  This paper describes the approach of the RescueRobots Freiburg team, which is a team of students from the University of Freiburg that originates from the former CS Freiburg team (RoboCupSoccer) and the ResQ Freiburg team (RoboCupRescue Simulation). Furthermore we introduce linkMAV, a micro aerial vehicle platform. Our approach covers RFID-based SLAM and exploration, autonomous detection of relevant 3D structures, visual odometry, and autonomous victim identification. Furthermore, we introduce a custom made 3D Laser Range Finder (LRF) and a novel mechanism for the active distribution of RFID tags.

• Christian Dornhege and Alexander Kleiner.
  Visual Odometry for Tracked Vehicles.
  In Proceedings of the IEEE International Workshop on Safety, Security and Rescue Robotics (SSRR 2006). 2006.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  Localization and mapping on autonomous robots typically requires a good pose estimate, which is hard to acquire if the vehicle is tracked. In this paper we describe a solution to the pose estimation problem by utilizing a consumer-quality camera and an Inertial Measurement Unit (IMU). The basic idea is to continuously track salient features with the KLT feature tracker over multiple images taken by the camera and to extract from the tracked features image vectors resulting from the robot's motion. Each image vector is taken for a voting that best explains the robot's motion. Image vectors vote according to a previously trained ^\m2102tile coding^\m2112 classificator that assigns to each possible image vector a translation probability. Our results show that the proposed single camera solution leads to sufficiently accurate pose estimates of the tracked vehicle.

• Alexander Kleiner, Michael Brenner, Tobias Braeuer, Christian Dornhege, Moritz Göbelbecker, Matthias Luber, Johann Prediger, Joerg Stueckler and Bernhard Nebel.
  Successful Search and Rescue in Simulated Disaster Areas.
  In Proceedings of the International RoboCup Symposium '05. Osaka, Japan 2005.
  (Show abstract) (Hide abstract) (PDF)
  

  RoboCupRescue Simulation is a large-scale multi-agent simulation of urban disasters where, in order to save lives and minimize damage, rescue teams must effectively cooperate despite sensing and communication limitations. This paper presents the comprehensive search and rescue approach of the ResQ Freiburg team, the winner in the RoboCupRescue Simulation league at RoboCup 2004. Specific contributions include the predictions of travel costs and civilian lifetime, the efficient coordination of an active disaster space exploration, as well as an any-time rescue sequence optimization based on a genetic algorithm. We compare the performances of our team and others in terms of their capability of extinguishing fires, freeing roads from debris, disaster space exploration, and civilian rescue. The evaluation is carried out with information extracted from simulation log files gathered during RoboCup 2004. Our results clearly explain the success of our team, and also confirm the scientific approaches proposed in this paper.

• Alexander Kleiner, Bastian Steder, Christian Dornhege, Daniel Hoefler, Daniel Meyer-Delius, Johann Prediger, Joerg Stueckler, Kolja Glogowski, Markus Thurner, Matthias Luber, Michael Schnell, Rainer Kuemmerle, Timothy Burk, Tobias Braeuer and Bernhard Nebel.
  RoboCupRescue - Robot League Team RescueRobots Freiburg (Germany), Team Description Paper.
  In CDROM Proceedings of the International RoboCup Symposium '05. Osaka, Japan 2005.
  (Show abstract) (Hide abstract) (PDF)
  

  This paper describes the approach of the RescueRobots Freiburg team. RescueRobots Freiburg is a team of students from the university of Freiburg, that originates from the former CS Freiburg team (RoboCupSoccer) and the ResQ Freiburg team (RoboCupRescue Simulation). Due to the high versatility of the RoboCupRescue competition we tackle the three arenas by a a twofold approach: On the one hand we want to introduce robust vehicles that can safely be teleoperated through rubble and building debris while constructing three-dimensional maps of the environment. On the other hand we want to introduce a team of autonomous robots that quickly explore a large terrain while building a two-dimensional map. This two solutions are particularly wellsuited for the red and yellow arena, respectively. Our solution for the orange arena will finally be decided between these two, depending on the capabilities of both approaches at the venue. In this paper, we introduce some preliminary results that we achieved so far from map building, localization, and autonomous victim identification. Furthermore we introduce a custom made 3D Laser Range Finder (LRF) and a novel mechanism for the active distribution of RFID tags. 1 Introduction RescueRobots Freiburg is a team of students from the university of Freiburg. The team originates from the former CS Freiburg team[6], which won three times the RoboCup world championship in the RoboCupSoccer F2000 league, and the ResQ Freiburg team[2], which won the last RoboCup world championship in the RoboCupRescue Simulation league. The team approach proposed in this paper is based on experiences gathered at RoboCup during the last six years. Due to the high versatility of the RoboCupRescue competition we tackle the three arenas by a twofold approach: On the one hand we want to introduce a vehicle that can safely be teleoperated through rubble and building debris while constructing threedimensional maps of the environment. On the other hand we want to introduce an autonomous team of robots that quickly explore a large terrain while building a twodimensional map. This two solutions are particularly well-suited for the red and yellow arena, respectively. Our solution for the orange arena will finally be decided between these two, depending on the capabilities of both approaches at the venue.

• Alexander Kleiner, Michael Brenner, Tobias Braeuer, Christian Dornhege, Moritz Göbelbecker, Matthias Luber, Johann Prediger and Joerg Stueckler.
  ResQ Freiburg: Team Description and Evaluation, Team Description Paper, Rescue Simulation League.
  In CDROM Proceedings of the International RoboCup Symposium '04. Lisbon, Portugal 2004.
  (PDF)
  

Christoph Dornheim

• Christoph Dornheim.
  Graph Embedding with Topological Cycle-Constraints.
  In J. Kratochvil (ed.), Proceedings of the 7th International Symposium on Graph Drawing (GD 1999). Stirin, Czech Republic 1999.
  (PS.GZ)
  

• Christoph Dornheim.
  Undecidability of Plane Polygonal Mereotopology.
  In A. G. Cohn, L. Schubert and S. C. Shapiro (eds.), Principles of Knowledge Representation and Reasoning, Proceedings of the 6th International Conference (KR'98). Trento, Italy 1998.
  (PS.GZ)
  

Thorsten Engesser

• Thorsten Engesser, Robert Mattmüller, Bernhard Nebel and Michael Thielscher.
  Game description language and dynamic epistemic logic compared.
  Artificial Intelligence 292. 2021.
  (Show abstract) (Hide abstract) (Online; DOI)
  

  Several different frameworks have been proposed to model and reason about knowledge in dynamic multi-agent settings, among them the logic-programming-based game description language GDL-III and dynamic epistemic logic (DEL). GDL-III and DEL have complementary strengths and weaknesses in terms of ease of modeling and simplicity of semantics. In this paper, we formally study the expressiveness of GDL-III vs. DEL. We clarify the commonalities and differences between those languages, demonstrate how to bridge the differences where possible, and identify large fragments of GDL-III and DEL that are equivalent in the sense that they can be used to encode games or planning tasks that admit the same legal action sequences. We prove the latter by providing translations between those fragments of GDL-III and DEL.

• Thorsten Engesser, Robert Mattmüller, Bernhard Nebel and Felicitas Ritter.
  Token-based Execution Semantics for Multi-Agent Epistemic Planning.
  In Proceedings of the 17th International Conference on Principles of Knowledge Representation and Reasoning (KR-2020), pp. 351-360. 2020.
  (Show abstract) (Hide abstract) (Online; PDF)
  

  Epistemic planning has been employed as a means to achieve implicit coordination in cooperative multi-agent systems where world knowledge is distributed between the agents, and agents plan and act individually. However, recent work has shown that even if all agents act with respect to plans that they consider optimal from their own subjective perspective, infinite executions can occur. In this paper, we analyze the idea of using a single token that can be passed around between the agents and which is used as a prerequisite for acting. We show that introducing such a token to any planning task will prevent the existence of infinite executions. We furthermore analyze the conditions under which solutions to a planning task are preserved under our tokenization.

• Thorsten Engesser and Tim Miller.
  Implicit Coordination Using FOND Planning.
  In Proceedings of the 34th AAAI Conference on Artificial Intelligence (AAAI-20). 2020.
  To appear.
  (Show abstract) (Hide abstract) (PDF)
  

  Epistemic planning can be used to achieve implicit coordination in cooperative multi-agent settings where knowledge and capabilities are distributed between the agents. In these scenarios, agents plan and act on their own without having to agree on a common plan or protocol beforehand. However, epistemic planning is undecidable in general. In this paper, we show how implicit coordination can be achieved in a simpler, propositional setting by using nondeterminism as a means to allow the agents to take the other agents' perspectives. We identify a decidable fragment of epistemic planning that allows for arbitrary initial state uncertainty and non-determinism, but where actions can never increase the uncertainty of the agents. We show that in this fragment, planning for implicit coordination can be reduced to a version of fully observable nondeterministic (FOND) planning and that it thus has the same computational complexity as FOND planning. We provide a small case study, modeling the problem of multi-agent path finding with destination uncertainty in FOND, to show that our approach can be successfully applied in practice.

• Daniel Reifsteck, Thorsten Engesser, Robert Mattmüller and Bernhard Nebel.
  Epistemic Multi-agent Planning Using Monte-Carlo Tree Search.
  In KI 2019: Advances in Artificial Intelligence - 42nd German Conference on AI, pp. 277-289. 2019.
  (Show abstract) (Hide abstract) (Online; DOI)
  

  Coordination in multi-agent systems with partial and non-uniform observability is a practically challenging problem. We use Monte-Carlo tree search as the basis of an implicitly coordinated epistemic planning algorithm which is capable of using the knowledge distributed among the agents to find solutions in problems even with a large branching factor. We use Dynamic Epistemic Logic to represent the knowledge and the actual situation as a state of the Monte-Carlo tree search, and epistemic planning to formalize the goals and actions of a problem. Further, we describe the required modifications of the Monte-Carlo tree search when searching over epistemic states, and make use of the cooperative card game Hanabi to test our planner on larger problems. We find that the approach scales to games with up to eight cards while maintaining high playing strength.

• Bernhard Nebel, Thomas Bolander, Thorsten Engesser, Robert Mattmüller and .
  Implicitly Coordinated Multi-Agent Path Finding under Destination Uncertainty: Success Guarantees and Computational Complexity (Extended Abstract).
  In Proceedings of International Joint Conference on Artificial Intelligence (IJCAI 2019), pp. 6372-6374. 2019.
  

• Thorsten Engesser and Tim Miller.
  Planning for Implicit Coordination using FOND.
  In Proceedings of the Workshop on Knowledge Engineering for Planning and Scheduling (KEPS19). 2019.
  Superseded by the AAAI-20 paper by the same authors.
  (Show abstract) (Hide abstract)
  

  Epistemic Planning can be used to achieve implicit coordination in cooperative multi-agent settings where knowledge and capabilities are distributed between the agents. In these scenarios, agents plan and act on their own without having to agree on a common plan or protocol beforehand. However, epistemic planning is undecidable in general. In this paper, we identify a decidable fragment of epistemic planning that allows for arbitrary initial state uncertainty and nondeterminism, but where actions can never increase the uncertainty of the agents. We show that in this fragment, planning with and without implicit coordination can be reduced to fully observable nondeterministic (FOND) planning and that it shares the same omputational complexity. We also provide a small case study, modeling the problem of multi-agent path finding with destination uncertainty in FOND, to show that our compilation approach can be successfully applied in practice.

• Thomas Bolander, Thorsten Engesser, Andreas Herzig, Robert Mattmüller and Bernhard Nebel.
  The Dynamic Logic of Policies and Contingent Planning.
  In Logics in Artificial Intelligence - 16th European Conference (JELIA-2019), pp. 659-674. 2019.
  (Show abstract) (Hide abstract) (Online; DOI)
  

  In classical deterministic planning, solutions to planning tasks are simply sequences of actions, but that is not sufficient for contingent plans in non-deterministic environments. Contingent plans are often expressed through policies that map states to actions. An alternative is to specify contingent plans as programs, e.g. in the syntax of Propositional Dynamic Logic (PDL). PDL is a logic for reasoning about programs with sequential composition, test and non-deterministic choice. However, as we show in the paper, none of the existing PDL modalities directly captures the notion of a solution to a planning task under non-determinism. We add a new modality to star-free PDL correctly capturing this notion. We prove the appropriateness of the new modality by showing how to translate back and forth between policies and PDL programs under the new modality. More precisely, we show how a policy solution to a planning task gives rise to a program solution expressed via the new modality, and vice versa. We also provide an axiomatisation of our PDL extension through reduction axioms into standard star-free PDL.

• Bernhard Nebel, Thomas Bolander, Thorsten Engesser and Robert Mattmüller.
  Implicitly Coordinated Multi-Agent Path Finding under Destination Uncertainty: Success Guarantees and Computational Complexity.
  Journal of Artificial Intelligence Research 64, pp. 497-527. 2019.
  (Show abstract) (Hide abstract) (PDF)
  

  In multi-agent path finding (MAPF), it is usually assumed that planning is performed centrally and that the destinations of the agents are common knowledge. We will drop both assumptions and analyze under which conditions it can be guaranteed that the agents reach their respective destinations using implicitly coordinated plans without communication. Furthermore, we will analyze what the computational costs associated with such a coordination regime are. As it turns out, guarantees can be given assuming that the agents are of a certain type. However, the implied computational costs are quite severe. In the distributed setting, we either have to solve a sequence of NP-complete problems or have to tolerate exponentially longer executions. In the setting with destination uncertainty, bounded plan existence becomes PSPACE-complete. This clearly demonstrates the value of communicating about plans before execution starts.

• Thomas Bolander, Thorsten Engesser, Robert Mattmüller and Bernhard Nebel.
  Better Eager Than Lazy? How Agent Types Impact the Successfulness of Implicit Coordination.
  In Proceedings of the Sixteenth Conference on Principles of Knowledge Representation and Reasoning (KR18), pp. 445-453. 2018.
  (Show abstract) (Hide abstract) (PDF)
  

  Epistemic planning can be used for decision making in multi-agent situations with distributed knowledge and capabilities. In recent work, we proposed a new notion of strong policies with implicit coordination. With this it is possible to solve planning tasks with joint goals from a single-agent perspective without the agents having to negotiate about and commit to a joint policy at plan time. We study how and under which circumstances the decentralized application of those policies leads to the desired outcome.

• Thorsten Engesser, Robert Mattmüller, Bernhard Nebel and Michael Thielscher.
  Game Description Language and Dynamic Epistemic Logic Compared.
  In Proceedings of International Joint Conference on Artificial Intelligence (IJCAI 2018), pp. 1795-1802. 2018.
  (Show abstract) (Hide abstract) (PDF)
  

  Several different frameworks have been proposed to model and reason about knowledge in dynamic multi-agent settings, among them the logic-programming-based game description language GDL-III, and dynamic epistemic logic (DEL), based on possible-worlds semantics. GDL-III and DEL have complementary strengths and weaknesses in terms of ease of modeling and simplicity of semantics. In this paper, we formally study the expressiveness of GDL-III vs. DEL. We clarify the commonalities and differences between those languages, demonstrate how to bridge the differences where possible, and identify large fragments of GDL-III and DEL that are equivalent in the sense that they can be used to encode games or planning tasks that admit the same legal action sequences. We prove the latter by providing compilations between those fragments of GDL-III and DEL.

• Thorsten Engesser, Thomas Bolander, Robert Mattmüller and Bernhard Nebel.
  Cooperative Epistemic Multi-Agent Planning for Implicit Coordination.
  In Ghosh, Sujata and Ramanujam and R. (eds.), Proceedings of the Ninth Workshop on Methods for Modalities (M4M 2017), pp. 75-90. 2017.
  (Show abstract) (Hide abstract) (Online; PDF) (BIB)
  

  Epistemic planning can be used for decision making in multi-agent situations with distributed knowledge and capabilities. Recently, Dynamic Epistemic Logic (DEL) has been shown to provide a very natural and expressive framework for epistemic planning. We extend the DEL-based epistemic planning framework to include perspective shifts, allowing us to define new notions of sequential and conditional planning with implicit coordination. With these, it is possible to solve planning tasks with joint goals in a decentralized manner without the agents having to negotiate about and commit to a joint policy at plan time. First we define the central planning notions and sketch the implementation of a planning system built on those notions. Afterwards we provide some case studies in order to evaluate the planner empirically and to show that the concept is useful for multi-agent systems in practice.

• Thomas Bolander, Thorsten Engesser, Robert Mattmüller and Bernhard Nebel.
  Better Eager Than Lazy? How Agent Types Impact the Successfulness of Implicit Coordination.
  In Proceedings of the ICAPS-2016 Workshop on Distributed and Multi-Agent Planning (DMAP 2016). 2016.
  Superseded by the KR18 paper by the same authors.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  Epistemic planning can be used for decision making in multi-agent situations with distributed knowledge and capabilities. Recent work proposed a new notion of strong policies with implicit coordination. With this it is possible to solve planning tasks with joint goals from a single-agent perspective without the agents having to negotiate about and commit to a joint policy at plan time. We study how and under which circumstances the decentralized application of those policies leads to the desired outcome.

• Thorsten Engesser, Thomas Bolander, Robert Mattmüller and Bernhard Nebel.
  Cooperative Epistemic Multi-Agent Planning With Implicit Coordination.
  In Proceedings of the ICAPS-2015 Workshop on Distributed and Multi-Agent Planning (DMAP 2015). 2015.
  Superseded by the M4M 2017 paper by the same authors.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  Epistemic Planning has been used to achieve ontic and epistemic control in multi-agent situations. We extend the formalism to include perspective shifts, allowing us to define a class of cooperative problems in which both action planning and execution is done in a purely distributed fashion, meaning coordination is only allowed implicitly by means of the available epistemic actions. While this approach can be fruitfully applied to model reasoning in some simple social situations, we also provide some benchmark applications to show that the concept is useful for multi-agent systems in practice.

Patrick Eyerich

• Patrick Eyerich and Malte Helmert.
  Stronger Abstraction Heuristics Through Perimeter Search.
  In Proceedings of the 23rd International Conference on Automated Planning and Scheduling (ICAPS13). 2013.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  Perimeter search is a bidirectional search algorithm consisting of two phases. In the first phase, a limited regression search computes the perimeter, a region which must necessarily be passed in every solution. In the second phase, a heuristic forward search finds an optimal plan from the initial state to the perimeter.

  The drawback of perimeter search is the need to compute heuristic estimates towards every state on the perimeter in the forward phase. We show that this limitation can be effectively overcome when using pattern database (PDB) heuristics in the forward phase.

  The combination of perimeter search and PDB heuristics has been considered previously by Felner and Ofek for solving combinatorial puzzles. They claimed that, based on theoretical considerations and experimental evidence, the use of perimeter search in this context offers "limited or no benefits". Our theoretical and experimental results show that this assessment should be revisited.

• Johannes Löhr, Patrick Eyerich, Stefan Winkler and Bernhard Nebel.
  Domain Predictive Control Under Uncertain Numerical State Information.
  In Proceedings of the 23rd International Conference on Automated Planning and Scheduling (ICAPS13). 2013.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  In planning, hybrid system states consisting of logical and numerical variables are usually assumed to be completely known. In particular, for numerical state variables full knowledge of their exact values is assumed. However, in real world applications states are results of noisy measurements and imperfect actuators. Therefore, a planned sequence of state transitions might fail to lead a hybrid system to the desired goal. We show how to propagate and reason about uncertain state information directly in the planning process, enabling hybrid systems to find plans that satisfy numerical goals with predefined confidence.

• Patrick Eyerich.
  Preferring Properly: Increasing Coverage while Maintaining Quality in Anytime Temporal Planning.
  In Proceedings of the 20th European Conference on Artificial Intelligence (ECAI 2012). 2012.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  Temporal Fast Downward (TFD) is a successful temporal planning system that is capable of dealing with numerical values. Rather than decoupling action selection from scheduling, it searches directly in the space of time-stamped states, an approach that has shown to produce plans of high quality at the price of coverage. To increase coverage, TFD incorporates deferred evaluation and preferred operators, two search techniques that usually decrease the number of heuristic calculations by a large amount. However, the current definition of preferred operators offers only limited guidance in problems where heuristic estimates are weak or where subgoals require the execution of mutex operators. In this paper, we present novel methods for refinement of this definition and show how to combine the diverse strengths of different sets of preferred operators using a restarting procedure incorporated into a multi-queue best-first search. These techniques improve TFD's coverage drastically and preserve the average solution quality, leading to a system that solves more problems than each of the competitors of the temporal satisficing track of IPC 2011 and clearly outperforms all of them in terms of IPC score.

• Thomas Keller and Patrick Eyerich.
  PROST: Probabilistic Planning Based on UCT.
  In Proceedings of the 22nd International Conference on Automated Planning and Scheduling (ICAPS 2012), pp. 119-127. 2012.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  We present PROST, a probabilistic planning system that is based on the UCT algorithm by Kocsis and Szepesvari (2006), which has been applied successfully to many areas of planning and acting under uncertainty. The objective of this paper is to show the application of UCT to domain-independent probabilistic planning, an area it had not been applied to before. We furthermore present several enhancements to the algorithm, including a method that is able to drastically reduce the branching factor by identifying superfluous actions. We show how search depth limitation leads to a more thoroughly investigated search space in parts that are influential on the quality of a policy, and present a sound and polynomially computable detection of reward locks, states that correspond to, e.g., dead ends or goals. We describe a general Q-value initialization for unvisited nodes in the search tree that circumvents the initial random walks inherent to UCT, and leads to a faster convergence on average. We demonstrate the significant influence of the enhancements by providing a comparison on the IPPC benchmark domains.

• Johannes Löhr, Patrick Eyerich, Thomas Keller and Bernhard Nebel.
  A Planning Based Framework for Controlling Hybrid Systems.
  In Proceedings of the 22nd International Conference on Automated Planning and Scheduling (ICAPS 2012). 2012.
  (Show abstract) (Hide abstract) (PDF)
  

  The control of dynamic systems, which aims to minimize the deviation of state variables from reference values in a contin- uous state space, is a central domain of cybernetics and con- trol theory. The objective of action planning is to find feasible state trajectories in a discrete state space from an initial state to a state satisfying the goal conditions, which in principle ad- dresses the same issue on a more abstract level. We combine these approaches to switch between dynamic system charac- teristics on the fly, and to generate control input sequences that affect both discrete and continuous state variables. Our approach (called Domain Predictive Control) is applicable to hybrid systems with linear dynamics and discretizable inputs.

• Patrick Eyerich.
  Preferring Properly: Increasing Coverage while Maintaining Quality in Anytime Temporal Planning.
  In Proceedings of the ICAPS-12 Workshop on Heuristics and Search for Domain Independent Planning (HSDIP 2012). 2012.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  Temporal Fast Downward (TFD) is a successful temporal planning system that is capable of dealing with numerical values. Rather than decoupling action selection from scheduling, it searches directly in the space of time-stamped states, an approach that has shown to produce plans of high quality at the price of coverage. To increase coverage, TFD incorporates deferred evaluation and preferred operators, two search techniques that usually decrease the number of heuristic calculations by a large amount. However, the current definition of preferred operators offers only limited guidance in problems where heuristic estimates are weak or where subgoals require the execution of mutex operators. In this paper, we present novel methods for refinement of this definition and show how to combine the diverse strengths of different sets of preferred operators using a restarting procedure incorporated into a multi-queue best-first search. These techniques improve TFD's coverage drastically and preserve the average solution quality, leading to a system that solves more problems than each of the competitors of the temporal satisficing track of IPC 2011 and clearly outperforms all of them in terms of IPC score.

• Thomas Keller and Patrick Eyerich.
  A Polynomial All Outcome Determinization for Probabilistic Planning.
  In Fahiem Bacchus, Carmel Domshlak, Stefan Edelkamp and Malte Helmert (eds.), Proceedings of the 21th International Conference on Automated Planning and Scheduling (ICAPS 2011), pp. 331-334. AAAI Press 2011.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  Most predominant approaches in probabilistic planning utilize techniques from the more thoroughly investigated field of classical planning by determinizing the problem at hand. In this paper, we present a method to map probabilistic operators to an equivalent set of probabilistic operators in a novel normal form, requiring polynomial time and space. From this, we directly derive a determinization which can be used for, \eg, replanning strategies incorporating a classical planning system. Unlike previously described all outcome determinizations, the number of deterministic operators is not exponentially but polynomially bounded in the number of parallel probabilistic effects, enabling the use of more sophisticated determinization-based techniques in the future.

• J. Benton, Patrick Eyerich and Subbarao Kambhampati.
  Enhancing Search for Satisficing Temporal Planning with Objective-driven Decisions.
  In Proceedings of the ICAPS-2011 Workshop on Heuristics for Domain-independent Planning, pp. 59-65. 2011.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  Heuristic best-first search techniques have recently enjoyed ever-increasing scalability in finding satisficing solutions to a variety of automated planning problems, and temporal planning is no different. Unfortunately, achieving efficient computational performance often comes at the price of clear guidance toward solution of high quality. This fact is sharp in the case of many best-first search temporal planners, who often use a node evaluation function that is mismatched with the objective function, reducing the likelihood that plans returned will have a short makespan but increasing search performance. To help mitigate matters, we introduce a method that works to progress search on actions declared ``useful'' according to makespan, even when the original search may ignore the makespan value of search nodes. We study this method and show that it increases over all plan quality in most of the benchmark domains from the temporal track of the 2008 International Planning Competition.

• Kai M. Wurm, Christian Dornhege, Patrick Eyerich, Cyrill Stachniss, Bernhard Nebel and Wolfram Burgard.
  Coordinated Exploration with Marsupial Teams of Robots using Temporal Symbolic Planning.
  In Proceedings of the 2010 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS 2010). 2010.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  The problem of autonomously exploring an environment with a team of robots received considerable attention in the past. However, there are relatively few approaches to coordinate teams of robots that are able to deploy and retrieve other robots. Efficiently coordinating the exploration with such marsupial robots requires advanced planning mechanisms that are able to consider symbolic deployment and retrieval actions. In this paper, we propose a novel approach for coordinating the exploration with marsupial robot teams. Our method integrates a temporal symbolic planner that explicitly considers deployment and retrieval actions with a traditional cost-based assignment procedure. Our approach has been implemented and evaluated in several simulated environments and with varying team sizes. The results demonstrate that our proposed method is able to coordinate marsupial teams of robots to efficiently explore unknown environments.

• Thomas Keller, Patrick Eyerich and Bernhard Nebel.
  Task Planning for an Autonomous Service Robot.
  In Rüdiger Dillmann, Jürgen Beyerer, Uwe Hanebeck and Tanja Schultz (eds.), Proceedings on the 33rd Annual German Conference on Artificial Intelligence (KI 2010), pp. 358-365. Springer-Verlag 2010.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  In the DESIRE project an autonomous robot capable of performing service tasks in a typical kitchen environment has been developed. The overall system consists of various loosely coupled subcomponents providing particular features like manipulating objects or recognizing and interacting with humans. To bring all these subcomponents together to act as monolithic system, a high-performance planning system has been implemented. In this paper, we present this system’s basic architecture and some advanced extensions necessary to cope with the various challenges arising in dynamic and uncertain environments like those a real world service robot is usually faced with.

• Patrick Eyerich, Thomas Keller and Malte Helmert.
  High-Quality Policies for the Canadian Traveler's Problem.
  In Maria Fox and David Poole (eds.), Proceedings of the Twenty-Fourth AAAI Conference on Artificial Intelligence (AAAI 2010), pp. 51-58. AAAI Press 2010.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  We consider the stochastic variant of the Canadian Traveler's Problem, a path planning problem where adverse weather can cause some roads to be untraversable. The agent does not initially know which roads can be used. However, it knows a probability distribution for the weather, and it can observe the status of roads incident to its location. The objective is to find a policy with low expected travel cost.

  We introduce and compare several algorithms for the stochastic CTP. Unlike the optimistic approach most commonly considered in the literature, the new approaches we propose take uncertainty into account explicitly. We show that this property enables them to generate policies of much higher quality than the optimistic one, both theoretically and experimentally.

• Patrick Eyerich, Thomas Keller and Malte Helmert.
  High-Quality Policies for the Canadian Traveler's Problem (Extended Abstract).
  In Ariel Felner and Nathan Sturtevant (eds.), Proceedings of the Third Annual Symposium on Combinatorial Search (SoCS 2010), pp. 147-148. AAAI Press 2010.
  Extended abstract of the AAAI paper by the same name.
  (PDF)
  

• J. Benton, Kartik Talamadupula, Patrick Eyerich, Robert Mattmüller and Subbarao Kambhampati.
  G-value Plateaus: A Challenge for Planning.
  In Ronen Brafman, Héctor Geffner, Jörg Hoffmann and Henry Kautz (eds.), Proceedings of the 20th International Conference on Automated Planning and Scheduling (ICAPS 2010), pp. 259-262. AAAI Press 2010.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  Recent years have seen the development of several scalable planners, many of which follow the string of successes found in using heuristic, best-first search methods. While this provides positive reinforcement for continuing work along these lines, fundamental problems arise when handling objectives whose value does not change with each search operation. An extreme case of this occurs when handling the objective of generating a temporal plan with short makespan. Typically used heuristic search methods assume strictly positive edge costs for their guarantees on completeness and optimality to hold, while the usual "fattening" and "advance time" steps of heuristic search algorithms for temporal planning have the potential for zero-cost edges, resulting in "g-value plateaus". In this paper we point out some underlying difficulties with using modern heuristic search methods for optimizing makespan and discuss how the presence of these problems contributes to the poor performance of makespan-optimizing heuristic search planners. To further illustrate this, we show empirical results on recent benchmarks using a planner made with makespan optimization in mind.

• Moritz Göbelbecker, Thomas Keller, Patrick Eyerich, Michael Brenner and Bernhard Nebel.
  Coming Up with Good Excuses: What To Do When No Plan Can be Found.
  In Ronen Brafman, Héctor Geffner, Jörg Hoffmann and Henry Kautz (eds.), Proceedings of the 20th International Conference on Automated Planning and Scheduling (ICAPS 2010), pp. 81-88. AAAI Press 2010.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  When using a planner-based agent architecture, many things can go wrong. First and foremost, an agent might fail to execute one of the planned actions for some reasons. Even more annoying, however, is a situation where the agent is incompetent, i.e., unable to come up with a plan. This might be due to the fact that there are principal reasons that prohibit a successful plan or simply because the task's description is incomplete or incorrect. In either case, an explanation for such a failure would be very helpful. We will address this problem and provide a formalization of coming up with excuses for not being able to find a plan. Based on that, we will present an algorithm that is able to find excuses and demonstrate that such excuses can be found in practical settings in reasonable time.

• Patrick Eyerich, Thomas Keller and Malte Helmert.
  High-Quality Policies for the Canadian Traveler's Problem.
  In Proceedings of the ICAPS-2010 Workshop on Planning and Scheduling Under Uncertainty. 2010.
  Superseded by the AAAI 2010 paper by the same name.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  We consider the stochastic variant of the Canadian Traveler's Problem, a path planning problem where adverse weather can cause some roads to be untraversable. The agent does not initially know which roads can be used. However, it knows a probability distribution for the weather, and it can observe the status of roads incident to its location. The objective is to find a policy with low expected travel cost.

  We introduce and compare several algorithms for the stochastic CTP. Unlike the optimistic approach most commonly considered in the literature, the new approaches we propose take uncertainty into account explicitly. We show that this property enables them to generate policies of much higher quality than the optimistic one, both theoretically and experimentally.

• Patrick Eyerich, Thomas Keller and Bernhard Nebel.
  Combining Action and Motion Planning via Semantic Attachments.
  In Proceedings of the Workshop on Combining Action and Motion Planning at ICAPS 2010 (CAMP 2010), p. 19. 2010.
  Extended Abstract.
  (PDF) (BIB)
  

• Christian Dornhege, Patrick Eyerich, Thomas Keller, Sebastian Trüg, Michael Brenner and Bernhard Nebel.
  Semantic Attachments for Domain-Independent Planning Systems.
  In Proceedings of the 19th International Conference on Automated Planning and Scheduling (ICAPS 2009), pp. 114-121. AAAI Press 2009.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  Solving real-world problems using symbolic planning often requires a simplified formulation of the original problem, since certain subproblems cannot be represented at all or only in a way leading to inefficiency. For example, manipulation planning may appear as a subproblem in a robotic planning context or a packing problem can be part of a logistics task. In this paper we propose an extension of PDDL for specifying semantic attachments. This allows the evaluation of grounded predicates as well as the change of fluents by externally specified functions. Furthermore, we describe a general schema of integrating semantic attachments into a forward-chaining planner and report on our experience of adding this extension to the planners FF and Temporal Fast Downward. Finally, we present some preliminary experiments using semantic attachments.

• Patrick Eyerich, Robert Mattmüller and Gabriele Röger.
  Using the Context-enhanced Additive Heuristic for Temporal and Numeric Planning.
  In Proceedings of the 19th International Conference on Automated Planning and Scheduling (ICAPS 2009), pp. 130-137. AAAI Press 2009.
  (Show abstract) (Hide abstract) (PDF) (slides; PDF) (BIB)
  

  Planning systems for real-world applications need the ability to handle concurrency and numeric fluents. Nevertheless, the predominant approach to cope with concurrency followed by the most successful participants in the latest International Planning Competitions (IPC) is still to find a sequential plan that is rescheduled in a post-processing step. We present Temporal Fast Downward (TFD), a planning system for temporal problems that is capable of finding low-makespan plans by performing a heuristic search in a temporal search space. We show how the context-enhanced additive heuristic can be successfully used for temporal planning and how it can be extended to numeric fluents. TFD often produces plans of high quality and, evaluated according to the rating scheme of the last IPC, outperforms all state-of-the-art temporal planning systems.

• Paul Plöger, Kai Pervölz, Christoph Mies, Patrick Eyerich, Michael Brenner and Bernhard Nebel.
  The DESIRE Service Robotics Initiative.
  Künstliche Intelligenz 08 (4), pp. 29-32. 2008.
  (Show abstract) (Hide abstract) (PDF)
  

  We present some advanced hardware units and an appropriate component based SW architecture for DESIRE. As an example we describe the integration of a enhanced AI task planner which allows for higher flexibility and dependability during complex task execution.

• Patrick Eyerich, Michael Brenner and Bernhard Nebel.
  On the Complexity of Planning Operator Subsumption.
  In Proceedings of the Eleventh International Conference on Principles of Knowledge Representation and Reasoning (KR 2008), pp. 518-527. AAAI Press 2008.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  Formal action models play a central role in several subfields of AI because they are used to model application domains, e.g., in automated planning. However, there are hitherto no automated methods for relating such domain models to each other, in particular for checking whether one is a specialization or generalization of the other. In this paper, we introduce two kinds of subsumption relations between operators, both of which are suitable for modeling and verifying hierarchies between actions and operators: applicability subsumption considers an action to be more general than another if the latter can be replaced by the first at each point in each sound sequence of actions; abstraction subsumption exploits relations between actions from an ontological point of view. For both kinds of subsumption, we prove complexity results for verifying operator subsumption in three important subclasses: The problems are NP-complete when the expressiveness of the operators is restricted to the well-known basic STRIPS formalism, Sigma_p_2-complete when we admit boolean logical operators and undecidable when the full power of the planning language ADL is permitted.

• Jens Claßen, Patrick Eyerich, Gerhard Lakemeyer and Bernhard Nebel.
  Towards an Integration of Golog and Planning.
  In Proceedings of the 20th International Joint Conference on Artificial Intelligence (IJCAI 2007), pp. 1846-1851. AAAI Press 2007.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  The action language Golog has been applied successfully to the control of robots, among other things. Perhaps its greatest advantage is that a user can write programs which constrain the search for an executable plan in a xible manner. However, when general planning is needed, Golog supports this only in principle, but does not measure up with state-of-the-art planners. In this paper we propose an integration of Golog and planning in the sense that planning problems, formulated as part of a Golog program, are solved by a modern planner during the execution of the program. Here we focus on the ADL subset of the plan language PDDL. First we show that the semantics of ADL can be understood as progression in the situation calculus, which underlies Golog, thus providing us with a correct embedding of ADL within Golog. We then show how Golog can be integrated with an existing ADL planner for closed-world initial databases and compare the performance of the resulting system with the original Golog.

• Patrick Eyerich, Bernhard Nebel, Gerhard Lakemeyer and Jens Classen.
  Golog and PDDL: What is the Relative Expressiveness?
  In Proceedings of the International Symposium on Practical Cognitive Agents and Robots (PCAR 2006), pp. 93-104. University of Western Australia Press 2006.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  Action formalisms such as GOLOG or FLUX have been developed primarily for representing and reasoning about change in a logical framework. For this reason, expressivity was the main goal in the development of these formalisms. In another line of research, efficiency of planning methods was the topmost goal resulting in the basic STRIPS language, which has only moderate expressivity. The planning language PDDL developed since 1998 is an extension of basic STRIPS with many expressive features. Now the interesting question is how PDDL compares to GOLOG or other action languages from an expressivity point of view. We will show that a GOLOG fragment, which we call Restricted Basic Action Theories, is as expressive as the ADL fragment of PDDL. To prove this equivalence we use the compilation framework. From a practical point of view, this result can be used for employing efficient planners inside a GOLOG interpreter.

Florian Geißer

• David Speck, Florian Geißer and Robert Mattmüller.
  When Perfect is not Good Enough: On the Search Behaviour of Symbolic Heuristic Search.
  In Proceedings of the 30th International Conference on Automated Planning and Scheduling (ICAPS 2020), pp. 263-271. 2020.
  (Show abstract) (Hide abstract) (PDF)
  

  Symbolic search has proven to be a competitive approach to cost-optimal planning, as it compactly represents sets of states by symbolic data structures. While heuristics for symbolic search exist, symbolic bidirectional blind search empirically outperforms its heuristic counterpart and is therefore the dominant search strategy. This prompts the question of why heuristics do not seem to pay off in symbolic search. As a first step in answering this question, we investigate the search behaviour of symbolic heuristic search by means of BDDA*. Previous work identified the partitioning of state sets according to their heuristic values as the main bottleneck. We theoretically and empirically evaluate the search behaviour of BDDA* and reveal another fundamental problem: we prove that the use of a heuristic does not always improve the search performance of BDDA*. In general, even the perfect heuristic can exponentially deteriorate search performance.

• Florian Geißer, David Speck and Thomas Keller.
  Trial-based Heuristic Tree Search for MDPs with Factored Action Spaces.
  In Proceedings of the 13th Annual Symposium on Combinatorial Search (SoCS 2020), pp. 38-47. 2020.
  (Show abstract) (Hide abstract) (PDF)
  

  MDPs with factored action spaces, i.e. where actions are described as assignments to a set of action variables, allow reasoning over action variables instead of action states, yet most algorithms only consider a grounded action representation. This includes algorithms that are instantiations of the trial-based heuristic tree search (THTS) framework, such as AO* or UCT. To be able to reason over factored action spaces, we propose a generalisation of THTS where nodes that branch over all applicable actions are replaced with subtrees that consist of nodes that represent the decision for a single action variable. We show that many THTS algorithms retain their theoretical properties under the generalised framework, and show how to approximate any state-action heuristic to a heuristic for partial action assignments. This allows to guide a UCT variant that is able to create exponentially fewer nodes than the same algorithm that considers ground actions. An empirical evaluation on the benchmark set of the probabilistic track of the latest International Planning Competition validates the benefits of the approach.

• Sumitra Corraya, Florian Geißer, David Speck and Robert Mattmüller.
  An Empirical Study of the Usefulness of State-Dependent Action Costs in Planning.
  In Proceedings of the 42nd German Conference on Artificial Intelligence (KI 2019), pp. 123-130. 2019.
  (Show abstract) (Hide abstract) (PDF) (PDF; Online)
  

  The vast majority of work in planning to date has focused on state-independent action costs. However, if a planning task features state-dependent costs, using a cost model with state-independent costs means either introducing a modeling error, or potentially sacrificing compactness of the model. In this paper, we investigate the conflicting priorities of modeling accuracy and compactness empirically, with a particular focus on the extent of the negative impact of reduced modeling accuracy on (a) the quality of the resulting plans, and (b) the search guidance provided by heuristics that are fed with inaccurate cost models. Our empirical results show that the plan suboptimality introduced by ignoring state-dependent costs can range, depending on the domain, from inexistent to several orders of magnitude. Furthermore, our results show that the impact on heuristic guidance additionally depends strongly on the heuristic that is used, the specifics of how exactly the costs are represented, and whether one is interested in heuristic accuracy, node expansions, or overall runtime savings.

• Florian Geißer, David Speck and Thomas Keller.
  An Analysis of the Probabilistic Track of the IPC 2018.
  In Proceedings of the ICAPS-2019 Workshop on the International Planning Competition (WIPC 2019), pp. 27-35. 2019.
  (Show abstract) (Hide abstract) (PDF)
  

  The International Planning Competition 2018 consisted of several tracks on classical, temporal and probabilistic planning. In this paper, we focus on the discrete MDP track of the probabilistic portion of the competition.

  We discuss the changes to the input language RDDL, which give rise to new challenges for planning systems, and analyze each of the eight competition domains separately and highlight unique properties. We demonstrate flaws of the used evaluation criterion, the IPC score, and discuss the need for optimal upper bounds. An evaluation of the three top-performers, including their post-competition versions, and a brief analysis of their performance highlights the strengths and weaknesses of the individual approaches.

• David Speck, Florian Geißer, Robert Mattmüller and Álvaro Torralba.
  Symbolic Planning with Axioms.
  In Proceedings of the 29th International Conference on Automated Planning and Scheduling (ICAPS 2019), pp. 464-572. 2019.
  (Show abstract) (Hide abstract) (PDF)
  

  Axioms are an extension for classical planning models that allow for modeling complex preconditions and goals exponentially more compactly. Although axioms were introduced in planning more than a decade ago, modern planning techniques rarely support axioms, especially in cost-optimal planning. Symbolic search is a popular and competitive optimal planning technique based on the manipulation of sets of states. In this work, we extend symbolic search algorithms to support axioms natively. We analyze different ways of encoding derived variables and axiom rules to evaluate them in a symbolic representation. We prove that all encodings are sound and complete, and empirically show that the presented approach outperforms the previous state of the art in cost-optimal classical planning with axioms.

• Florian Geißer.
  On planning with state-dependent action costs.
  FreiDok plus 2018.
  Dissertation Thesis.
  (PDF)
  

• David Speck, Florian Geißer and Robert Mattmüller.
  Symbolic Planning with Edge-Valued Multi-Valued Decision Diagrams.
  In Proceedings of the 28th International Conference on Automated Planning and Scheduling (ICAPS 2018). 2018.
  (Show abstract) (Hide abstract) (PDF) (technical report with proofs; PDF)
  

  Symbolic representations have attracted significant attention in optimal planning. Binary Decision Diagrams (BDDs) form the basis for symbolic search algorithms. Closely related are Algebraic Decision Diagrams (ADDs), used to represent heuristic functions.

  Also, progress was made in dealing with models that take state-dependent action costs into account. Here, costs are represented as Edge-valued Multi-valued Decision Diagrams (EVMDDs), which can be exponentially more compact than the corresponding ADD representation. However, they were not yet considered for symbolic planning.

  In this work, we study EVMDD-based symbolic search for optimal planning. We define EVMDD-based representations of state sets and transition relations, and show how to compute the necessary operations required for EVMDD-A*. This EVMDD-based version of symbolic A* generalizes its BDD variant, and allows to solve planning tasks with state-dependent action costs.

  We prove theoretically that our approach is sound, complete and optimal. Additionally, we present an empirical analysis comparing EVMDD-A* to BDD-A* and explicit A* search. Our results underscore the usefulness of symbolic approaches and the feasibility of dealing with models that go beyond unit costs.

• Florian Geißer and David Speck.
  PROST-DD - Utilizing Symbolic Classical Planning in THTS.
  In The 6th International Probabilistic Planning Competition (IPPC 2018), pp. 13-16. 2018.
  (Show abstract) (Hide abstract) (PDF)
  

  We describe PROST-DD, our submission to the International Probabilistic Planning Competition 2018. Like its predecessor PROST, which already participated with success at the previous IPPC, PROST-DD is based on the trial-based heuristic tree search framework and applies the UCT* algorithm. The novelty of our submission is the heuristic used to initialize newly encountered decision nodes. We apply an iterative symbolic backward planning approach based on the determinized task. Similarly to the SPUDD approach and recent work in symbolic planning with state-dependent action costs, we encode costs and reachability of states in a single decision diagram. During initialization, these diagrams are then used to query a state for its estimated expected reward. One benefit of this heuristic is that we can optionally interweave the standard heuristic of PROST, the IDS heuristic.

• David Speck, Florian Geißer and Robert Mattmüller.
  SYMPLE: Symbolic Planning based on EVMDDs.
  In The 9th International Planning Competition (IPC 2018), pp. 82-85. 2018.
  (Show abstract) (Hide abstract) (PDF)
  

  SYMPLE is a classical planner which performs bidirectional symbolic search. Symbolic search has proven to be a useful approach to optimal classical planning and is usually based on Binary Decision Diagrams (BDDs). Our approach is based on an alternative data structure called Edge-valued Multi-valued Decision Diagrams (EVMDDs), which have some structural advantages over BDDs.

• Robert Mattmüller, Florian Geißer, Benedict Wright and Bernhard Nebel.
  On the Relationship Between State-Dependent Action Costs and Conditional Effects in Planning.
  In Proceedings of the Thirty-Second AAAI Conference on Artificial Intelligence (AAAI 2018). 2018.
  (Show abstract) (Hide abstract) (PDF)
  

  When planning for tasks that feature both state-dependent action costs and conditional effects using relaxation heuristics, the following problem appears: handling costs and effects separately leads to worse-than-necessary heuristic values, since we may get the more useful effect at the lower cost by choosing different values of a relaxed variable when determining relaxed costs and relaxed active effects.

  In this paper, we show how this issue can be avoided by representing state-dependent costs and conditional effects uniformly, both as edge-valued multi-valued decision diagrams (EVMDDs) over different sets of edge values, and then working with their product diagram. We develop a theory of EVMDDs that is general enough to encompass state-dependent action costs, conditional effects, and even their combination.

  We define relaxed effect semantics in the presence of state-dependent action costs and conditional effects, and describe how this semantics can be efficiently computed using product EVMDDs. This will form the foundation for informative relaxation heuristics in the setting with state-dependent costs and conditional effects combined.

• Robert Mattmüller, Florian Geißer, Benedict Wright and Bernhard Nebel.
  On the Relationship Between State-Dependent Action Costs and Conditional Effects in Planning.
  In Proceedings of the 9th Workshop on Heuristics and Search for Domain-Independent Planning (HSDIP 2017). 2017.
  Superseded by the AAAI 2018 paper by the same name.
  (Show abstract) (Hide abstract) (PDF)
  

  When planning for tasks that feature both state-dependent action costs and conditional effects using relaxation heuristics, the following problem appears: handling costs and effects separately leads to worse-than-necessary heuristic values, since we may get the more useful effect at the lower cost by choosing different values of a relaxed variable when determining relaxed costs and relaxed active effects.

  In this paper, we show how this issue can be avoided by representing state-dependent costs and conditional effects uniformly, both as edge-valued multi-valued decision diagrams (EVMDDs) over different sets of edge values, and then working with their product diagram. We develop a theory of EVMDDs that is general enough to encompass state-dependent action costs, conditional effects, and even their combination.

  We define relaxed effect semantics in the presence of state-dependent action costs and conditional effects, and describe how this semantics can be efficiently computed using product EVMDDs. This will form the foundation for informative relaxation heuristics in the setting with state-dependent costs and conditional effects combined.

• Dali Sun, Florian Geißer and Bernhard Nebel.
  Towards Effective Localization in Dynamic Environments.
  In Proceedings of the IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS 2016). 2016.
  (Show abstract) (Hide abstract) (PDF)
  

  Localization in dynamic environments is still a challenging problem in robotics – especially if rapid and large changes occur irregularly. Inspired by SLAM algorithms, our Bayesian approach to this so-called dynamic localization problem divides it into a localization problem and a mapping problem, respectively. To tackle the localization problem we use a particle filter, coupled with a distance filter and a scan matching method, which achieves a more robust localization against dynamic obstacles. For the mapping problem we use an extended sensor model which results in an effective and precise map update effect. We compare our approach against other localization methods and evaluate the impact the map update effect has on the localization in dynamic environments.

• Thomas Keller, Florian Pommerening, Jendrik Seipp, Florian Geißer and Robert Mattmüller.
  State-dependent Cost Partitionings for Cartesian Abstractions in Classical Planning (Extended Abstract).
  In Proceedings of the 39th German Conference on Artificial Intelligence (KI 2016). 2016.
  (Show abstract) (Hide abstract) (PDF)
  

  Abstraction heuristics are a popular method to guide optimal search algorithms in classical planning. Cost partitionings allow to sum heuristic estimates admissibly by partitioning action costs among the abstractions. We introduce state-dependent cost partitionings which take context information of actions into account, and show that an optimal state-dependent cost partitioning dominates its state-independent counterpart. We demonstrate the potential of state-dependent cost partitionings with a state-dependent variant of the recently proposed saturated cost partitioning, and show that it can sometimes improve not only over its state-independent counterpart, but even over the optimal state-independent cost partitioning.

• Thomas Keller, Florian Pommerening, Jendrik Seipp, Florian Geißer and Robert Mattmüller.
  State-dependent Cost Partitionings for Cartesian Abstractions in Classical Planning.
  In Proceedings of the 25th International Joint Conference on Artificial Intelligence (IJCAI 2016). 2016.
  (Show abstract) (Hide abstract) (PDF)
  

  Abstraction heuristics are a popular method to guide optimal search algorithms in classical planning. Cost partitionings allow to sum heuristic estimates admissibly by distributing action costs among the heuristics. We introduce state-dependent cost partitionings which take context information of actions into account, and show that an optimal state-dependent cost partitioning dominates its state-independent counterpart. We demonstrate the potential of our idea with a state-dependent variant of the recently proposed saturated cost partitioning, and show that it has the potential to improve not only over its state-independent counterpart, but even over the optimal state-independent cost partitioning. Our empirical results give evidence that ignoring the context of actions in the computation of a cost partitioning leads to a significant loss of information.

• Florian Geißer, Thomas Keller and Robert Mattmüller.
  Abstractions for Planning with State-Dependent Action Costs.
  In Proceedings of the 26th International Conference on Automated Planning and Scheduling (ICAPS 2016). 2016.
  (Show abstract) (Hide abstract) (PDF)
  

  Extending the classical planning formalism with state-dependent action costs (SDAC) allows an up to exponentially more compact task encoding. Recent work proposed to use edge-valued multi-valued decision diagrams (EVMDDs) to represent cost functions, which allows to automatically detect and exhibit structure in cost functions and to make heuristic estimators accurately reflect SDAC. However, so far only the inadmissible additive heuristic has been considered in this context. In this paper, we define informative admissible abstraction heuristics which enable optimal planning with SDAC. We discuss how abstract cost values can be extracted from EVMDDs that represent concrete cost functions without adjusting them to the selected abstraction. Our theoretical analysis shows that this is efficiently possible for abstractions that are Cartesian or coarser. We adapt the counterexample-guided abstraction refinement approach to derive such abstractions. An empirical evaluation of the resulting heuristic shows that highly accurate values can be computed quickly.

• Florian Geißer, Thomas Keller and Robert Mattmüller.
  Delete Relaxations for Planning with State-Dependent Action Costs.
  In Proceedings of the 24th International Joint Conference on Artificial Intelligence (IJCAI 2015). 2015.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  Most work in planning focuses on tasks with state-independent or even uniform action costs. However, supporting state-dependent action costs admits a more compact representation of many tasks. We investigate how to solve such tasks using heuristic search, with a focus on delete-relaxation heuristics. We first define a generalization of the additive heuristic to such tasks and then discuss different ways of computing it via compilations to tasks with state-independent action costs and more directly by modifying the relaxed planning graph. We evaluate these approaches theoretically and present an implementation of the generalized additive heuristic for planning with state-dependent action costs. To our knowledge, this gives rise to the first approach able to handle even the hardest instances of the combinatorial Academic Advising domain from the International Probabilistic Planning Competition (IPPC) 2014.

• Florian Geißer, Thomas Keller and Robert Mattmüller.
  Delete Relaxations for Planning with State-Dependent Action Costs.
  In Proceedings of the 8th Annual Symposium on Combinatorial Search (SoCS 2015). 2015.
  Extended abstract of the IJCAI 2015 paper by the same name.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  Supporting state-dependent action costs in planning admits a more compact representation of many tasks. We generalize the additive heuristic and compute it by embedding decision-diagram representations of action cost functions into the RPG. We give a theoretical evaluation and present an implementation of the generalized additive heuristic. This allows us to handle even the hardest instances of the combinatorial Academic Advising domain from the IPPC 2014.

• Thomas Keller and Florian Geißer.
  Better Be Lucky Than Good: Exceeding Expectations in MDP Evaluation.
  In Proceedings of the 29th AAAI Conference on Artificial Intelligence (AAAI 2015). AAAI Press 2015.
  Erratum: On page 7, we mention that the results at IPPC would have differed by "-0.09", "+0.04" and "+0.05", which should read "-0.009", "+0.004" and "+0.005" instead.
  (Show abstract) (Hide abstract) (PDF)
  

  We introduce the MDP-Evaluation Stopping Problem, the optimization problem faced by participants of the International Probabilistic Planning Competition 2014 that focus on their own performance. It can be constructed as a meta-MDP where actions correspond to the application of a policy on a base-MDP, which is intractable in practice. Our theoretical analysis reveals that there are tractable special cases where the problem can be reduced to an optimal stopping problem. We derive approximate strategies of high quality by relaxing the general problem to an optimal stopping problem, and show both theoretically and experimentally that it not only pays off to pursue luck in the execution of the optimal policy, but that there are even cases where it is better to be lucky than good as the execution of a suboptimal base policy is part of an optimal strategy in the meta-MDP.

• Florian Geißer, Thomas Keller and Robert Mattmüller.
  Past, Present, and Future: An Optimal Online Algorithm for Single-Player GDL-II Games.
  In Proceedings of the 21st European Conference on Artificial Intelligence (ECAI 2014), pp. 357-362. 2014.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  In General Game Playing, a player receives the rules of an unknown game and attempts to maximize his expected reward. Since 2011, the GDL-II rule language extension allows the formulation of nondeterministic and partially observable games. In this paper, we present an algorithm for such games, with a focus on the single-player case. Conceptually, at each stage, the proposed Norns algorithm distinguishes between the past, present and future steps of the game. More specifically, a belief state tree is used to simulate a potential past that leads to a present that is consistent with received observations. Unlike other related methods, our method is asymptotically optimal. Moreover, augmenting the belief state tree with iteratively improved probabilities speeds up the process over time significantly. As this allows a true picture of the present, we additionally present an optimal version of the well-known UCT algorithm for partially observable single-player games. Instead of performing hindsight optimization on a simplified, fully observable tree, the true future is simulated on an action-observation tree that takes partial observability into account. The expected reward estimates of applicable actions converge towards the true expected rewards even for moves that are only used to gather information. We prove that our algorithm is asymptotically optimal for single-player games and POMDPs and support our claim with an empirical evaluation.

Jens-Steffen Gutmann

• Thilo Weigel, Jens-Steffen Gutmann, Markus Dietl, Alexander Kleiner and Bernhard Nebel.
  CS Freiburg: Coordinating Robots for Successful Soccer Playing.
  IEEE Transactions on Robotics and Automation 18 (5), pp. 685-699. 2002.
  (Show abstract) (Hide abstract) (PDF)
  

  Robotic soccer is a challenging research domain because many different research areas have to be addressed in order to create a successful team of robot players. This paper presents the CS Freiburg team, the winner in the middle size league at RoboCup 1998, 2000 and 2001. The paper focuses on multi-agent coordination for both perception and action. The contributions of this work are new methods for tracking ball and players observed by multiple robots, team coordination methods for strategic team formation and dynamic role assignment, a rich set of basic skills allowing to respond to large range of situations in an appropriate way, an action selection method based on behavior networks as well as a method to learn the skills and their selection. As demonstrated by evaluations of the different methods and by the success of the team, these methods permit the creation of a multi-robot group, which is able to play soccer successfully. In addition, the developed methods promise to advance the state of the art in the multi-robot field.

• Markus Dietl, Jens-Steffen Gutmann and Bernhard Nebel.
  Cooperative Sensing in Dynamic Environments.
  In Proceedings of the IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS-2001). 2001.
  (Show abstract) (Hide abstract) (PDF)
  

  This work presents methods for tracking objects from noisy and unreliable data taken by a team of robots. We develop a multi­object tracking algorithm based on Kalman filtering and a single­object tracking method involving a combination of Kalman filtering and Markov localization for outlier detection. We apply these methods in the context of robot soccer for robots participating in the middle­size league and compare them to a simple averaging method. Results including situations from real competition games are presented.

• Markus Dietl, Jens-Steffen Gutmann and Bernhard Nebel.
  CS Freiburg: Global View by Cooperative Sensing.
  In International RoboCup Symposium 2001. 2001.
  (Show abstract) (Hide abstract) (PDF)
  

  Global vision systems as found in the small size league are prohibited in the middle size league. This paper presents methods for creating a global view of the world by cooperative sensing of a team of robots. We develop a multi­object tracking algorithm based on Kalman filtering and a single­object tracking method involving a combination of Kalman filtering and Markov localization for outlier detection. We apply these methods for robots participating in the middle­size league and compare them to a simple averaging method. Results including situations from real competition games are presented.

• Jens-Steffen Gutmann, Thilo Weigel and Bernhard Nebel.
  A Fast, Accurate, and Robust Method for Self-Localization in Polygonal Environments Using Laser-Range-Finders.
  Advanced Robotics 14 (8), pp. 651-668. 2001.
  (Show abstract) (Hide abstract) (PDF)
  

  Self-localization is important in almost all robotic tasks. For playing an aesthetic and effective game of robotic soccer, self-localization is a necessary prerequisite. When we designed our robotic soccer team for participating in robotic soccer competitions, it turned out that all existing approaches did not meet our requirements of being fast, accurate, and robust. For this reason, we developed a new method, which is presented and analyzed in this paper. This method is one of the key components and is probably one of the explanations for the success of our team in national and international competitions. We present also experimental evidence that our method outperforms other self-localization methods in the RoboCup environment.

• Thilo Weigel, Willi Auerbach, Markus Dietl, Burkhard Dümler, Jens-Steffen Gutmann, Kornel Marko, Klaus Müller, Bernhard Nebel, Boris Szerbakowski and Maximilian Thiel.
  CS Freiburg: Doing the Right Thing in a Group.
  In P. Stone, G. Kraetzschmar and T. Balch (eds.), RoboCup 2000: Robot Soccer World Cup IV, pp. 52-63. Springer-Verlag 2001.
  (Show abstract) (Hide abstract) (PDF)
  

  The success of CS Freiburg at RoboCup 2000 can be attributed to an effective cooperation between players based on sophisticated soccer skills and a robust and accurate self-localization method. In this paper, we present our multi-agent coordination approach for both, action and perception, and our rich set of basic skills which allow to respond to a large range of situations in an appropriate way. Furthermore our action selection method based on an extension to behavior networks is described. Results including statistics from CS Freiburg final games at RoboCup 2000 are presented.

• Thilo Weigel, Alexander Kleiner, Florian Diesch, Markus Dietl, Jens-Steffen Gutmann, Bernhard Nebel, Patrick Stiegeler and Boris Szerbakowski.
  CS Freiburg 2001.
  In International RoboCup Symposium 2001. 2001.
  (Show abstract) (Hide abstract) (PDF)
  

  The CS Freiburg team has become F2000 champion the third time in the history of RoboCup. The success of our team can probably be attributed to its robust sensor interpretation and its team play. In this paper, we will focus on new developments in our vision system, in our path planner, and in the cooperation component.

• Thilo Weigel, Jens-Steffen Gutmann, Bernhard Nebel, Klaus Müller and Markus Dietl.
  CS Freiburg: Sophisticated Skills and Effective Cooperation.
  In Proc. European Control Conference (ECC-01). Porto, Portugal 2001.
  (Show abstract) (Hide abstract) (PDF)
  

  The success of CS Freiburg at RoboCup 2000 can be attributed to a robust and accurate perception approach and an effec­ tive cooperation between players based on sophisticated soc­ cer skills. In this paper, we present our multi­agent coordi­ nation approach for both, action and perception, and our rich set of basic skills which allow to respond to a large range of situations in an appropriate way. Furthermore, our action se­ lection method based on an extension to behavior networks is described. Results including statistics from CS Freiburg final games at RoboCup 2000 are presented.

• Jens-Steffen Gutmann, Wolfgang Hatzack, Immanuel Herrmann, Bernhard Nebel, Frank Rittinger, Augustinus Topor and Thilo Weigel.
  The CS Freiburg Team: Playing Robotic Soccer Based on an Explicit World Model.
  AI Magazine 21 (1), pp. 37-46. 2000.
  (Show abstract) (Hide abstract) (preliminary version; PDF)
  

  Robotic soccer is an ideal task to demonstrate new techniques and to explore new problems. Moreover, problems and solutions can be easily communicated because soccer is a well-known game. Our intention in building a robotic soccer team and participating in RoboCup'98 was, first of all, to demonstrate the usefulness of the self-localization methods we have developed. Secondly, we wanted to show that playing soccer based on an explicit world model is much more effective than other methods. Thirdly, we intended to explore the problem of building and maintaining a global team world model. As has been demonstrated by the performance of our team, we were successful on the first two points. Moreover, robotic soccer gave us the opportunity to study problems in distributed, cooperative sensing.

• Jens-Steffen Gutmann, Bernhard Nebel and Christian Reetz.
  CS Freiburg: Architektur und Aktionsauswahl im Roboterfuball.
  In Proc. AMS-2000. 2000.
  (Show abstract) (Hide abstract) (PDF)
  

  Roboterfußball ist ein wissenschaftliches anspruchsvolles Forschungsproblem, das erfordert, Probleme aus den Bereichen Robotik, Künstliche Intelligenz und Multi-Agenten-Systeme zu lösen und die Lösungen in einem System zu integrieren, um ein erfolgreiches Roboterfußballteam zu kreieren. In diesem Papier beschreiben wir die Schlüsselkomponenten des CS Freiburg Teams. Dabei fokussieren wir auf die Selbstlokalisation und Objekterkennungsmethoden und die Integration aller Information in ein globales Weltmodell. Basierend auf diesem Weltmodell werden dann Aktionsselektion, Pfadplanung und Kooperation realisiert. Das resultierende System ist äußerst erfolgreich und hat bisher lediglich ein Spiel in einem Wettbewerb verloren.

• Jens-Steffen Gutmann, Thilo Weigel and Bernhard Nebel.
  Fast, Accurate, and Robust Self-Localization in Polygonal Environments.
  In Proceedings of the IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS '99). Kyongju, Korea 1999.
  (Show abstract) (Hide abstract) (preliminary version; PDF)
  

  Self-localization is important in almost all robotic tasks. For playing an aesthetic and effective game of robotic soccer, self-localization is a necessary prerequisite. When we designed our robotic soccer team for RoboCup'98, it turned out that all existing approaches did not meet our requirements of being fast, accurate, and robust. For this reason, we developed a new method, which is presented and analyzed in this paper. We additionally present experimental evidence that our method outperforms other methods in the RoboCup environment.

• Jens-Steffen Gutmann, Wolfgang Hatzack, Immanuel Herrmann, Bernhard Nebel, Frank Rittinger, Augustinus Topor, Thilo Weigel and Bruno Welsch.
  The CS Freiburg Robotic Soccer Team: Reliable Self-Localization, Multirobot Sensor Integration, and Basic Soccer Skills.
  In M. Asada (ed.), RoboCup-98: Robot Soccer World Cup II, pp. 93-108. Springer-Verlag, Berlin, Heidelberg, New York 1999.
  (Show abstract) (Hide abstract) (PDF)
  

  Robotic soccer is a challenging research domain because problems in robotics, artificial intelligence, multi-agent systems and real-time reasoning have to be solved in order to create a successful team of robotic soccer players. In this paper, we describe the key components of the CS Freiburg team. We focus on the self-localization and object recognition method based on using laser range finders and the integration of all this information into a global world model. Using the explicit model of the environment built by these components, we have implemented path planning, simple ball handling skills and basic multi-agent cooperation. The resulting system is a very successful robotic soccer team, which has not lost any game yet.

• Jens-Steffen Gutmann, Wolfgang Hatzack, Immanuel Herrmann, Bernhard Nebel, Frank Rittinger, Augustinus Topor and Thilo Weigel.
  Reliable Self-Localization, Multirobot Sensor Integration, Accurate Path-Planning and Basic Soccer Skills: Playing an Effective Game of Robotic Soccer.
  In Nineth International Conference on Advanced Robotics (ICAR 1999). 1999.
  (Show abstract) (Hide abstract) (PDF)
  

  Robotic soccer is a challenging research domain because problems in robotics, artificial intelligence, multi-agent systems and real-time reasoning have to be solved in order to create a successful team of robotic soccer players. In this paper, we describe the key components of the CS Freiburg team. We focus on the self-localization and object recognition method based on using laser range finders and the integration of all this information into a global world model. Using the explicit model of the environment built by these components, we have implemented path planning, simple ball handling skills and basic multi-agent cooperation. The resulting system is a very successful robotic soccer team, which has not lost any official game yet.

• Bernhard Nebel, Jens-Steffen Gutmann and Wolfgang Hatzack.
  The CS Freiburg '99 Team.
  In Third International Workshop on RoboCup. 1999.
  (Show abstract) (Hide abstract) (PDF)
  

  Based on the design of the CS Freiburg team, which participated successfully in Robocup'98, we developed a new team of robotic soccer players. While the hardware components and software architecture remained mainly unchanged, we invested some effort to improve the sensor data gathering and interpretation, the tactical components and the behavior-based control module. The main goal is to enable the players to act in a truly cooperative style which leads, for instance, to passing the ball from one player to another.

• Jens-Steffen Gutmann, Wolfram Burgard, Dieter Fox and Kurt Konolige.
  An Experimental Comparison of Localization Methods.
  In International Conference on Intelligent Robots and Systems (IROS 98). Victoria, Canada 1998.
  (PS.GZ)
  

• Bernhard Nebel, Wolfgang Hatzack, Thilo Weigel, Jens-Steffen Gutmann, Immanuel Herrmann, Frank Rittinger and Augustinus Topor.
  CS Freiburg's Participation at RoboCup'98: The World Champions in Robotic Soccer.
  AI Communications 11, pp. 243-248. 1998.
  (Show abstract) (Hide abstract) (PDF)
  

  Robotic soccer is a challenging research domain that can be used to explore new problems and to demonstrate new techniques. We participated in RoboCup'98 in order to explore the problems of cooperation in multi-robot-systems and to demonstrate our self-localization techniques based on laser range finders. In this paper we sketch the main technical points of our team, give a description of the process of developing our team before and during the competition, and describe how we viewed the competition in general.

• Sebastian Thrun, Jens-Steffen Gutmann, Dieter Fox, Wolfram Burgard and Benjamin J. Kuipers.
  Integrating Topological and Metric Maps for Mobile Robot Navigation: A Statistical Approach.
  In Proceedings of the 15th National Conference on Artificial Intelligence (AAAI-98). 1998.
  (PS.GZ)
  

• Jens-Steffen Gutmann and Bernhard Nebel.
  Navigation mobiler Roboter mit Laserscans.
  In Autonome Mobile Systeme 1997 (AMS'97), pp. 36-47. Springer-Verlag 1997.
  (Show abstract) (Hide abstract) (PDF)
  

  Es wird ein Verfahren zur Erstellung einer topologischen Karte aus Laserscandaten für die Navigation mobiler Roboter beschrieben. Aus einem Satz sich korrekt überdeckender 360 Grad Scans wird ein Sichtbarkeitsgraph erstellt, wobei Knoten Scanpositionen und Kanten die relative Anzahl gemeinsamer Scanpunkte (genannt Sichtbarkeit) repräsentieren. Aus der Sichtbarkeit und der Distanz der Scanpositionen wird eine subjektive Wahrscheinlichkeit für die Befahrbarkeit zwischen den Scanpositionen berechnet. Durch Annahme von Unabhängigkeit der berechneten Wahrscheinlichkeiten wird mittels uniformer Kostensuche ein möglichst kurzer und sicher befahrbarer Pfad bestimmt. Das Verfahren wurde auf einem Pioneer-1-Roboter mit SICK-Laserscanner implementiert und erprobt. Für die Navigation zu jedem Zwischenziel entlang des Pfades wurde ein gitterbasierter lokaler Wegeplaner verwendet. Dadurch konnte ein hoher Grad an Robustheit erlangt werden. Das System ist in der Lage unvorhergesehenen Hindernissen auszuweichen, nicht passierbare Wege zu erkennen und alternative Wege zu finden.

• Jens-Steffen Gutmann and Christian Schlegel.
  AMOS: Comparison of Scan Matching Approaches for Self-Localization in Indoor Environments.
  In Proceedings of the First Euromicro Workshop on Advanced Mobile Robots (EUROBOT '96), pp. 61-68. 1996.
  (PS.GZ)
  

Moritz Göbelbecker

• Johannes Aldinger, Robert Mattmüller and Moritz Göbelbecker.
  Complexity Issues of Interval Relaxed Numeric Planning.
  In KI 2015: Advances in Artificial Intelligence (KI 2015). 2015.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  Automated planning is computationally hard even in its most basic form as STRIPS planning. We are interested in numeric planning with instantaneous actions, a problem that is not decidable in general. Relaxation is an approach to simplifying complex problems in order to obtain guidance in the original problem. We present a relaxation approach with intervals for numeric planning and show that plan existence can be decided in polynomial time for tasks where dependencies between numeric effects are acyclic.

• Johannes Aldinger, Robert Mattmüller and Moritz Göbelbecker.
  Complexity Issues of Interval Relaxed Numeric Planning.
  In Proceedings of the ICAPS-2015 Workshop on Heuristic and Search for Domain-Independent Planning (HSDIP 2015). 2015.
  Superseded by the KI 2015 paper of the same name..
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  Automated planning is a hard problem even in its most basic form as STRIPS planning. We are interested in numeric planning tasks with instantaneous actions, a problem which is not even decidable in general. Relaxation is an approach to simplifying complex problems in order to obtain guidance in the original problem. In this paper we present a relaxation approach with intervals for numeric planning and discuss the arising complexity issues.

• Alper Aydemir, Andrzej Pronobis, Moritz Göbelbecker and Patric Jensfelt.
  Active Visual Object Search in Unknown Environments Using Uncertain Semantics.
  IEEE Transactions on Robotics 29 (4), pp. 986-1002. 2013.
  

• Alper Aydemir, Moritz Göbelbecker, Andrzej Pronobis, Kristoffer Sjöö and Patric Jensfelt.
  Plan-based Object Search and Exploration Using Semantic Spatial Knowledge in the Real World.
  In Proceedings of the 5th European Conference on Mobile Robotics (ECMR 2011). 2011.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  In this paper we present a principled planner based approach to the active visual object search problem in unknown environments. We make use of a hierarchical planner that combines the strength of decision theory and heuristics. Furthermore, our object search approach leverages on the conceptual spatial knowledge in the form of object cooccurences and semantic place categorisation. A hierarchical model for representing object locations is presented with which the planner is able to perform indirect search. Finally we present real world experiments to show the feasibility of the approach.

• Moritz Göbelbecker, Charles Gretton and Richard W. Dearden.
  A Switching Planner for Combined Task and Observation Planning.
  In Proceedings of the 25th AAAI Conference on Artificial Intelligence (AAAI 2011). 2011.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  From an automated planning perspective the problem of practical mobile robot control in realistic environments poses many important and contrary challenges. On the one hand, the planning process must be lightweight, robust, and timely. Over the lifetime of the robot it must always respond quickly with new plans that accommodate exogenous events, changing objectives, and the underlying unpredictability of the environment. On the other hand, in order to promote efficient behaviours the planning process must perform computationally expensive reasoning about contingencies and possible revisions of subjective beliefs according to quantitatively modelled uncertainty in acting and sensing. Towards addressing these challenges, we develop a continual planning approach that switches between using a fast satisficing ``classical'' planner, to decide on the overall strategy, and decision-theoretic planning to solve small abstract subproblems where deeper consideration of the sensing model is both practical, and can significantly impact overall performance. We evaluate our approach in large problems from a realistic robot exploration domain.

• Moritz Göbelbecker, Alper Aydemir, Andrzej Pronobis, Kristoffer Sjöö and Patric Jensfelt.
  A Planning Approach to Active Visual Search in Large Environments.
  In Proceedings of the AAAI-11 Workshop on Automated Action Planning for Autonomous Mobile Robots (PAMR). 2011.
  Workshop version of the ECMR11 paper "Plan-based Object Search and Exploration Using Semantic Spatial Knowledge in the Real World".
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  In this paper we present a principled planner based approach to the active visual object search problem in unknown environments. We make use of a hierarchical planner that combines the strength of decision theory and heuristics. Furthermore, our object search approach leverages on the conceptual spatial knowledge in the form of object co-occurrences and semantic place categorisation. A hierarchical model for representing object locations is presented with which the planner is able to perform indirect search. Finally we present real world experiments to show the feasibility of the approach.

• Moritz Göbelbecker, Charles Gretton and Richard W. Dearden.
  A Switching Planner for Combined Task and Observation Planning.
  In Electronic Proceedings of the Workshop on Decision Making in Partially Observable, Uncertain Worlds: Exploring Insights from Multiple Communities at the Twenty-Second International Join Conference on Artificial Intelligence (DMPOUW 2011). 2011.
  Workshop version of the AAAI11 paper of the same title..
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  From an automated planning perspective the problem of practical mobile robot control in realistic environments poses many important and contrary challenges. On the one hand, the planning process must be lightweight, robust, and timely. Over the lifetime of the robot it must always respond quickly with new plans that accommodate exogenous events, changing objectives, and the underlying unpredictability of the environment. On the other hand, in order to promote efficient behaviours the planning process must perform computationally expensive reasoning about contingencies and possible revisions of subjective beliefs according to quantitatively modelled uncertainty in acting and sensing. Towards addressing these challenges, we develop a continual planning approach that switches between using a fast satisficing ``classical'' planner, to decide on the overall strategy, and decision-theoretic planning to solve small abstract subproblems where deeper consideration of the sensing model is both practical, and can significantly impact overall performance. We evaluate our approach in large problems from a realistic robot exploration domain.

• Marc Hanheide, Charles Gretton, Richard Dearden, Nick Hawes, Jeremy Wyatt, Andrzej Pronobis, Alper Aydemir, Moritz Göbelbecker and Hendrik Zender.
  Exploiting Probabilistic Knowledge under Uncertain Sensing for Efficient Robot Behaviour.
  In Proceedings of the Twenty-Second International Joint Conference on Artificial Intelligence (IJCAI 2011). 2011.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  Robots must perform tasks efficiently and reliably while acting under uncertainty. One way to achieve efficiency is to give the robot common-sense knowledge about the structure of the world. Reliable robot behaviour can be achieved by modelling the uncertainty in the world probabilistically. We present a robot system that combines these two approaches and demonstrate the improvements in efficiency and reliability that result. Our first contribution is a probabilistic relational model integrating common-sense knowledge about the world in general, with observations of a particular environment. Our second contribution is a continual planning system which is able to plan in the large problems posed by that model, by automatically switching between decision-theoretic and classical procedures. We evaluate our system on object search tasks in two different real-world indoor environments. By reasoning about the trade-offs between possible courses of action with different informational effects, and exploiting the cues and general structures of those environments, our robot is able to consistently demonstrate efficient and reliable goal-directed behaviour.

• D. Skočaj, M. Kristan, A. Leonardis, M. Mahnič, A. Vrečko, M. Janíček, G.-J. M. Kruijff, P. Lison, M. Zillich, C. Gretton, M. Hanheide and Moritz Göbelbecker.
  A system approach to interactive learning of visual concepts.
  In Tenth International Conference on Epigenetic Robotics (EPIROB 2010). 2010.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  In this work we present a system and underlying mechanisms for continuous learning of visual concepts in dialogue with a human.

• Marc Hanheide, Nick Hawes, Jeremy Wyatt, Moritz Göbelbecker, Michael Brenner, Kristoffer Sjöö, Alper Aydemir, Patric Jensfelt, Hendrik Zender and Geert-Jan Kruijff.
  A Framework for Goal Generation and Management.
  In Proceedings of the AAAI Workshop on Goal-Directed Autonomy. 2010.
  (Show abstract) (Hide abstract) (BIB)
  

  Goal-directed behaviour is often viewed as an essential char- acteristic of an intelligent system, but mechanisms to generate and manage goals are often overlooked. This paper addresses this by presenting a framework for autonomous goal gener- ation and selection. The framework has been implemented as part of an intelligent mobile robot capable of exploring unknown space and determining the category of rooms au- tonomously. We demonstrate the efficacy of our approach by comparing the performance of two versions of our inte- grated system: one with the framework, the other without. This investigation leads us conclude that such a framework is desirable for an integrated intelligent system because it re- duces the complexity of the problems that must be solved by other behaviour-generation mechanisms, it makes goal- directed behaviour more robust in the face of a dynamic and unpredictable environments, and it provides an entry point for domain-specific knowledge in a more general system.

• Moritz Göbelbecker, Thomas Keller, Patrick Eyerich, Michael Brenner and Bernhard Nebel.
  Coming Up with Good Excuses: What To Do When No Plan Can be Found.
  In Ronen Brafman, Héctor Geffner, Jörg Hoffmann and Henry Kautz (eds.), Proceedings of the 20th International Conference on Automated Planning and Scheduling (ICAPS 2010), pp. 81-88. AAAI Press 2010.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  When using a planner-based agent architecture, many things can go wrong. First and foremost, an agent might fail to execute one of the planned actions for some reasons. Even more annoying, however, is a situation where the agent is incompetent, i.e., unable to come up with a plan. This might be due to the fact that there are principal reasons that prohibit a successful plan or simply because the task's description is incomplete or incorrect. In either case, an explanation for such a failure would be very helpful. We will address this problem and provide a formalization of coming up with excuses for not being able to find a plan. Based on that, we will present an algorithm that is able to find excuses and demonstrate that such excuses can be found in practical settings in reasonable time.

• Nick Hawes, Marc Hanheide, Kristoffer Sjöö, Alper Aydemir, Patric Jensfelt, Moritz Göbelbecker, Michael Brenner, Hendrik Zender, Pierre Lison, Ivana Kruijff-Korbayov, Geert-Jan M. Kruijff and Michael Zillich.
  Dora The Explorer: A Motivated Robot.
  In Proc. of 9th Int. Conf. on Autonomous Agents and Multiagent Systems (AAMAS 2010). 2010.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  Dora the Explorer is a mobile robot with a sense of curios- ity and a drive to explore its world. Given an incomplete tour of an indoor environment, Dora is driven by internal motivations to probe the gaps in her spatial knowledge. She actively explores regions of space which she hasn't previously visited but which she expects will lead her to further unex- plored space. She will also attempt to determine the cate- gories of rooms through active visual search for functionally important objects, and through ontology-driven inference on the results of this search.

• Moritz Göbelbecker and Christian Dornhege.
  Realistic Cities in Simulated Environments - An Open Street Map to Robocup Rescue Converter.
  In Online-Proceedings of the Fourth International Workshop on Synthetic Simulation and Robotics to Mitigate Earthquake Disaster (SRMED 2009). 2009.
  (Show abstract) (Hide abstract) (PDF)
  

  A general problem when developing large scale disaster simulation environments is to acquire GIS data. In this work, we tackle the problem of map generation from public sources. Usually the major problem is not only the data conversion itself, but to get access to the data at all. We solve this problem by using the website OpenStreetMap.org, that provides mapping data for the whole world in a wiki-style concept, as our source of data, thus being able to generate maps for almost any city. The data is converted to the format required by the Robocup Rescue Simulation System, enabling simulations on various real-world scenarios.

• Alexander Kleiner, Michael Brenner, Tobias Braeuer, Christian Dornhege, Moritz Göbelbecker, Matthias Luber, Johann Prediger, Joerg Stueckler and Bernhard Nebel.
  Successful Search and Rescue in Simulated Disaster Areas.
  In Proceedings of the International RoboCup Symposium '05. Osaka, Japan 2005.
  (Show abstract) (Hide abstract) (PDF)
  

  RoboCupRescue Simulation is a large-scale multi-agent simulation of urban disasters where, in order to save lives and minimize damage, rescue teams must effectively cooperate despite sensing and communication limitations. This paper presents the comprehensive search and rescue approach of the ResQ Freiburg team, the winner in the RoboCupRescue Simulation league at RoboCup 2004. Specific contributions include the predictions of travel costs and civilian lifetime, the efficient coordination of an active disaster space exploration, as well as an any-time rescue sequence optimization based on a genetic algorithm. We compare the performances of our team and others in terms of their capability of extinguishing fires, freeing roads from debris, disaster space exploration, and civilian rescue. The evaluation is carried out with information extracted from simulation log files gathered during RoboCup 2004. Our results clearly explain the success of our team, and also confirm the scientific approaches proposed in this paper.

• Alexander Kleiner, Michael Brenner, Tobias Braeuer, Christian Dornhege, Moritz Göbelbecker, Matthias Luber, Johann Prediger and Joerg Stueckler.
  ResQ Freiburg: Team Description and Evaluation, Team Description Paper, Rescue Simulation League.
  In CDROM Proceedings of the International RoboCup Symposium '04. Lisbon, Portugal 2004.
  (PDF)
  

Wolfgang Hatzack

• Wolfgang Hatzack and Bernhard Nebel.
  Solving the Operational Traffic Control Problem.
  In A. Cesta and D. Borrajo (eds.), Proceedings of the 6th European Conference on Planning (ECP 2001). 2001.
  (Show abstract) (Hide abstract) (PDF)
  

  The operational traffic control problem comes up in a number of different contexts. It involves the coordinated movement of a set of vehicles and has by and large the flavor of a scheduling problem. In trying to apply scheduling techniques to the problem, one notes that this is a job-shop scheduling problem with blocking, a type of scheduling problem that is quite unusual. In particular, we will highlight a condition necessary to guarantee that job-shop schedules can be executed in the presences of the blocking constraint. Based on the insight that the traffic problem is a scheduling problem, we can derive the computational complexity of the operational traffic control problem and can design some algorithms to deal with this problem. In particular, we will specify a very simple method that works well in fast-time simulation contexts.

• Jens-Steffen Gutmann, Wolfgang Hatzack, Immanuel Herrmann, Bernhard Nebel, Frank Rittinger, Augustinus Topor and Thilo Weigel.
  The CS Freiburg Team: Playing Robotic Soccer Based on an Explicit World Model.
  AI Magazine 21 (1), pp. 37-46. 2000.
  (Show abstract) (Hide abstract) (preliminary version; PDF)
  

  Robotic soccer is an ideal task to demonstrate new techniques and to explore new problems. Moreover, problems and solutions can be easily communicated because soccer is a well-known game. Our intention in building a robotic soccer team and participating in RoboCup'98 was, first of all, to demonstrate the usefulness of the self-localization methods we have developed. Secondly, we wanted to show that playing soccer based on an explicit world model is much more effective than other methods. Thirdly, we intended to explore the problem of building and maintaining a global team world model. As has been demonstrated by the performance of our team, we were successful on the first two points. Moreover, robotic soccer gave us the opportunity to study problems in distributed, cooperative sensing.

• Jens-Steffen Gutmann, Wolfgang Hatzack, Immanuel Herrmann, Bernhard Nebel, Frank Rittinger, Augustinus Topor, Thilo Weigel and Bruno Welsch.
  The CS Freiburg Robotic Soccer Team: Reliable Self-Localization, Multirobot Sensor Integration, and Basic Soccer Skills.
  In M. Asada (ed.), RoboCup-98: Robot Soccer World Cup II, pp. 93-108. Springer-Verlag, Berlin, Heidelberg, New York 1999.
  (Show abstract) (Hide abstract) (PDF)
  

  Robotic soccer is a challenging research domain because problems in robotics, artificial intelligence, multi-agent systems and real-time reasoning have to be solved in order to create a successful team of robotic soccer players. In this paper, we describe the key components of the CS Freiburg team. We focus on the self-localization and object recognition method based on using laser range finders and the integration of all this information into a global world model. Using the explicit model of the environment built by these components, we have implemented path planning, simple ball handling skills and basic multi-agent cooperation. The resulting system is a very successful robotic soccer team, which has not lost any game yet.

• Jens-Steffen Gutmann, Wolfgang Hatzack, Immanuel Herrmann, Bernhard Nebel, Frank Rittinger, Augustinus Topor and Thilo Weigel.
  Reliable Self-Localization, Multirobot Sensor Integration, Accurate Path-Planning and Basic Soccer Skills: Playing an Effective Game of Robotic Soccer.
  In Nineth International Conference on Advanced Robotics (ICAR 1999). 1999.
  (Show abstract) (Hide abstract) (PDF)
  

  Robotic soccer is a challenging research domain because problems in robotics, artificial intelligence, multi-agent systems and real-time reasoning have to be solved in order to create a successful team of robotic soccer players. In this paper, we describe the key components of the CS Freiburg team. We focus on the self-localization and object recognition method based on using laser range finders and the integration of all this information into a global world model. Using the explicit model of the environment built by these components, we have implemented path planning, simple ball handling skills and basic multi-agent cooperation. The resulting system is a very successful robotic soccer team, which has not lost any official game yet.

• Bernhard Nebel, Jens-Steffen Gutmann and Wolfgang Hatzack.
  The CS Freiburg '99 Team.
  In Third International Workshop on RoboCup. 1999.
  (Show abstract) (Hide abstract) (PDF)
  

  Based on the design of the CS Freiburg team, which participated successfully in Robocup'98, we developed a new team of robotic soccer players. While the hardware components and software architecture remained mainly unchanged, we invested some effort to improve the sensor data gathering and interpretation, the tactical components and the behavior-based control module. The main goal is to enable the players to act in a truly cooperative style which leads, for instance, to passing the ball from one player to another.

• Bernhard Nebel, Wolfgang Hatzack, Thilo Weigel, Jens-Steffen Gutmann, Immanuel Herrmann, Frank Rittinger and Augustinus Topor.
  CS Freiburg's Participation at RoboCup'98: The World Champions in Robotic Soccer.
  AI Communications 11, pp. 243-248. 1998.
  (Show abstract) (Hide abstract) (PDF)
  

  Robotic soccer is a challenging research domain that can be used to explore new problems and to demonstrate new techniques. We participated in RoboCup'98 in order to explore the problems of cooperation in multi-robot-systems and to demonstrate our self-localization techniques based on laser range finders. In this paper we sketch the main technical points of our team, give a description of the process of developing our team before and during the competition, and describe how we viewed the competition in general.

Malte Helmert

• Thomas Keller and Malte Helmert.
  Trial-based Heuristic Tree Search for Finite Horizon MDPs.
  In Proceedings of the 1st Multidisciplinary Conference on Reinforcement Learning and Decision Making (RLDM 2013), pp. 101-105. 2013.
  Extended abstract of the ICAPS 2013 paper by the same name.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  Dynamic programming is a well-known approach for solving MDPs. In large state spaces, asynchronous versions like Real-Time Dynamic Programming (RTDP) have been applied successfully. If unfolded into equivalent trees, Monte-Carlo Tree Search algorithms are a valid alternative. UCT, the most popular representative, obtains good anytime behavior by guiding the search towards promising areas of the search tree and supporting non-admissible heuristics. The global Heuristic Search algorithm AO* finds optimal solutions for MDPs that can be represented as acyclic AND/OR graphs. Despite the differences, these approaches actually have much in common. We present the Trial-based Heuristic Tree Search (THTS) framework that subsumes these approaches and distinguishes them based on only five ingredients: heuristic function, backup function, action selection, outcome selection, and trial length. We describe the ingredients that model RTDP, AO* and UCT within this framework, and use THTS to combine attributes of these algorithms step by step in order to derive novel algorithms with superior theoretical properties. We merge Full Bellman and Monte-Carlo backup functions to Partial Bellman backups, and gain a function that both allows partial updates and a procedure that labels states when they are solved. DP-UCT combines attributes and theoretical properties from RTDP and UCT even though it differs from the latter only in the used Partial Bellman backups. Our main algorithm, UCT* adds a limited trial length to DP-UCT to inherit the global search behavior of AO*, which ensures that parts of the state space that are closer to the root are investigated more thoroughly. The experimental evaluation shows that both DP-UCT and UCT* are not only superior to UCT, but also outperform P ROST, the winner of the International Probabilistic Planning Competition (IPPC) 2011 on the benchmarks of IPPC 2011.

• Martin Wehrle, Malte Helmert, Dr. Yusra Alkhazraji and Robert Mattmüller.
  The Relative Pruning Power of Strong Stubborn Sets and Expansion Core.
  In Proceedings of the 23rd International Conference on Automated Planning and Scheduling (ICAPS13). 2013.
  (Show abstract) (Hide abstract) (PDF)
  

  In the last years, pruning techniques based on partial order reduction have found increasing attention in the planning community. One recent result is that the expansion core method is a special case of the strong stubborn sets method proposed in model checking. However, it is still an open question if there exist efficiently computable strong stubborn sets with strictly higher pruning power than expansion core. In this paper, we prove that the pruning power of strong stubborn sets is strictly higher than the pruning power of expansion core even for a straight-forward instantiation of strong stubborn sets. This instantiation is as efficiently computable as expansion core. Hence, our theoretical results suggest that strong stubborn sets should be preferred to expansion core. Our empirical evaluation on all optimal benchmarks from the international planning competitions up to 2011 supports the theoretical results.

• Patrick Eyerich and Malte Helmert.
  Stronger Abstraction Heuristics Through Perimeter Search.
  In Proceedings of the 23rd International Conference on Automated Planning and Scheduling (ICAPS13). 2013.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  Perimeter search is a bidirectional search algorithm consisting of two phases. In the first phase, a limited regression search computes the perimeter, a region which must necessarily be passed in every solution. In the second phase, a heuristic forward search finds an optimal plan from the initial state to the perimeter.

  The drawback of perimeter search is the need to compute heuristic estimates towards every state on the perimeter in the forward phase. We show that this limitation can be effectively overcome when using pattern database (PDB) heuristics in the forward phase.

  The combination of perimeter search and PDB heuristics has been considered previously by Felner and Ofek for solving combinatorial puzzles. They claimed that, based on theoretical considerations and experimental evidence, the use of perimeter search in this context offers "limited or no benefits". Our theoretical and experimental results show that this assessment should be revisited.

• Thomas Keller and Malte Helmert.
  Trial-based Heuristic Tree Search for Finite Horizon MDPs.
  In Proceedings of the 23rd International Conference on Automated Planning and Scheduling (ICAPS 2013), pp. 135-143. 2013.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  Dynamic programming is a well-known approach for solving MDPs. In large state spaces, asynchronous versions like Real-Time Dynamic Programming have been applied successfully. If unfolded into equivalent trees, Monte-Carlo Tree Search algorithms are a valid alternative. UCT, the most popular representative, obtains good anytime behavior by guiding the search towards promising areas of the search tree. The Heuristic Search algorithm AO∗ finds optimal solutions for MDPs that can be represented as acyclic AND/OR graphs.

  We introduce a common framework, Trial-based Heuristic Tree Search, that subsumes these approaches and distinguishes them based on five ingredients: heuristic function, backup function, action selection, outcome selection, and trial length. Using this framework, we describe three new algorithms which mix these ingredients in novel ways in an attempt to combine their different strengths. Our evaluation shows that two of our algorithms not only provide superior theoretical properties to UCT, but also outperform state-of-the-art approaches experimentally.

• Dr. Yusra Alkhazraji, Martin Wehrle, Robert Mattmüller and Malte Helmert.
  A Stubborn Set Algorithm for Optimal Planning.
  In Proceedings of the 20th European Conference on Artificial Intelligence (ECAI 2012). 2012.
  (Show abstract) (Hide abstract) (PDF)
  

  We adapt a partial order reduction technique based on stubborn sets, originally proposed for detecting dead ends in Petri Nets, to the setting of optimal planning. We demonstrate that stubborn sets can provide significant state space reductions on standard planning benchmarks, outperforming the expansion core method.

• Silvan Sievers, Manuela Ortlieb and Malte Helmert.
  Efficient Implementation of Pattern Database Heuristics for Classical Planning.
  In Proceedings of the Fifth Annual Symposium on Combinatorial Search (SoCS 2012), pp. 105-111. AAAI Press 2012.
  (Show abstract) (Hide abstract) (PDF)
  

  Despite their general success in the heuristic search community, pattern database (PDB) heuristics have, until very recently, not been used by the most successful classical planning systems. We describe a new efficient implementation of pattern database heuristics within the Fast Downward planner. A planning system using this implementation is competitive with the state of the art in optimal planning, significantly improving over results from the previous best PDB heuristic implementation in planning.

• Raz Nissim, Jörg Hoffmann and Malte Helmert.
  Computing Perfect Heuristics in Polynomial Time: On Bisimulation and Merge-and-Shrink Abstractions in Optimal Planning.
  In Proceedings of the Twenty-Second International Joint Conference on Artificial Intelligence (IJCAI 2011), pp. 1983-1990. 2011.
  Erratum: In Section 7, we introduce greedy bisimulation as only respecting the bisimulation property for transitions (s, l, s') where sd(s) <= sd(s'). The implementation we evaluate in Section 8 is actually even more greedy than that, only respecting transitions where sd(s) < sd(s'). Using the definition from Section 7 leads to a strategy that behaves very similarly to the strategies using regular (non-greedy) bisimulation on these benchmarks..
  (Show abstract) (Hide abstract) (PDF)
  

  A* with admissible heuristics is a very successful approach to optimal planning. But how to derive such heuristics automatically? Merge-and-shrink abstraction (M&S) is a general approach to heuristic design whose key advantage is its capability to make very fine-grained choices in defining abstractions. However, little is known about how to actually make these choices. We address this via the well-known notion of bisimulation. When aggregating only bisimilar states, M&s yields a perfect heuristic. Alas, bisimulations are exponentially large even in trivial domains. We show how to apply label reduction -- not distinguishing between certain groups of operators -- without incurring any information loss, while potentially reducing bisimulation size exponentially. In several benchmark domains, the resulting algorithm computes perfect heuristics in polynomial time. Empirically, we show that approximating variants of this algorithm improve the state of the art in M&S heuristics. In particular, a hybrid of two such variants is competitive with the leading heuristic LM-cut.

• Carmel Domshlak, Malte Helmert, Erez Karpas, Emil Keyder, Silvia Richter, Gabriele Röger, Jendrik Seipp and Matthias Westphal.
  BJOLP: The Big Joint Optimal Landmarks Planner (planner abstract).
  In Seventh International Planning Competition (IPC 2011), Deterministic Part, pp. 91-95. 2011.
  (Show abstract) (Hide abstract) (PDF)
  

  BJOLP, The Big Joint Optimal Landmarks Planner uses landmarks to derive an admissible heuristic, which is then used to guide a search for a cost-optimal plan. In this paper we review landmarks and describe how they can be used to derive an admissible heuristic. We conclude with presenting the BJOLP planner.

• Silvia Richter, Matthias Westphal and Malte Helmert.
  LAMA 2008 and 2011 (planner abstract).
  In Seventh International Planning Competition (IPC 2011), Deterministic Part, pp. 50-54. 2011.
  (Show abstract) (Hide abstract) (PDF)
  

  LAMA is a propositional planning system based on heuristic search with landmarks. This paper describes two versions of LAMA that were entered into the 2011 International Planning Competition: the original LAMA as developed for the 2008 competition and a new re-implementation of LAMA that uses the latest version of the Fast Downward Planning Framework.

  Landmarks are propositions that must be true in every solution of a planning task. LAMA uses a heuristic derived from landmarks in conjunction with the well-known FF heuristic. LAMA builds on the Fast Downward Planning System using non-binary (but finite domain) state variables and multi-heuristic search. A weighted A* search is used with iteratively decreasing weights, so that the planner continues to search for plans of better quality until the search is terminated. LAMA combines cost-to-goal and distance-to-goal estimates with the aim of finding good solutions using reasonable runtime.

• Malte Helmert and Carmel Domshlak.
  LM-Cut: Optimal Planning with the Landmark-Cut Heuristic (planner abstract).
  In Seventh International Planning Competition (IPC 2011), Deterministic Part, pp. 103-105. 2011.
  (Show abstract) (Hide abstract) (PDF)
  

  The LM-Cut planner uses the landmark-cut heuristic, introduced by the authors in 2009, within a standard A* progression search framework to find optimal sequential plans for STRIPS-style planning tasks. This short paper recapitulates the main ideas surrounding the landmark-cut heuristic and provides pointers for further reading.

• Raz Nissim, Jörg Hoffmann and Malte Helmert.
  The Merge-and-Shrink Planner: Bisimulation-based Abstraction for Optimal Planning (planner abstract).
  In Seventh International Planning Competition (IPC 2011), Deterministic Part, pp. 106-107. 2011.
  (Show abstract) (Hide abstract) (PDF)
  

  Merge-and-shrink abstraction is a general approach to heuristic design whose key advantage is its capability to make very fine-grained choices in defining abstractions. The Merge-and-shrink planner uses two different strategies for making these choices, both based on the well-known notion of bisimulation. The resulting heuristics are used in two sequential runs of A* search.

• Carmel Domshlak, Malte Helmert, Erez Karpas and Shaul Markovitch.
  The SelMax Planner: Online Learning for Speeding up Optimal Planning (planner abstract).
  In Seventh International Planning Competition (IPC 2011), Deterministic Part, pp. 108-112. 2011.
  (Show abstract) (Hide abstract) (PDF)
  

  The SelMax planner combines two state-of-the-art admissible heuristics using an online learning approach. In this paper we describe the online learning approach employed by SelMax, briefly review the Fast Downward framework, and describe the SelMax planner.

• Malte Helmert, Gabriele Röger, Jendrik Seipp, Erez Karpas, Jörg Hoffmann, Emil Keyder, Raz Nissim, Silvia Richter and Matthias Westphal.
  Fast Downward Stone Soup (planner abstract).
  In Seventh International Planning Competition (IPC 2011), Deterministic Part, pp. 38-45. 2011.
  (Show abstract) (Hide abstract) (PDF)
  

  Fast Downward Stone Soup is a sequential portfolio planner that uses various heuristics and search algorithms that have been implemented in the Fast Downward planning system.

  We present a simple general method for concocting "planner soups", sequential portfolios of planning algorithms, and describe the actual recipes used for Fast Downward Stone Soup in the sequential optimization and sequential satisficing tracks of IPC 2011.

• Chris Fawcett, Malte Helmert, Holger Hoos, Erez Karpas, Gabriele Röger and Jendrik Seipp.
  FD-Autotune: Automated Configuration of Fast Downward (planner abstract).
  In Seventh International Planning Competition (IPC 2011), Deterministic Part, pp. 31-37. 2011.
  (Show abstract) (Hide abstract) (PDF)
  

  The FD-Autotune submissions for the IPC-2011 sequential tracks consist of three instantiations of the latest, highly parametric version of the Fast Downward Planning Framework. These instantiations have been automatically configured for performance on a wide range of planning domains, using the well-known ParamILS configurator. Two of the instantiations were entered into the sequential satisficing track and one into the sequential optimising track. We describe how the extremely large configuration space of Fast Downward was restricted to a subspace that, although still very large, can be managed by state-of-the-art automated configuration procedures, and how ParamILS was then used to obtain performance-optimised configurations.

• Chris Fawcett, Malte Helmert, Holger Hoos, Erez Karpas, Gabriele Röger and Jendrik Seipp.
  FD-Autotune: Domain-Specific Configuration using Fast Downward (planner abstract).
  In Seventh International Planning Competition (IPC 2011), Planning and Learning Part. 2011.
  (Show abstract) (Hide abstract) (PDF)
  

  The FD-Autotune learning planning system is based on the idea of domain-specific configuration of the latest, highly parametric version of the Fast Downward Planning Framework by means of a generic automated algorithm configuration procedure. We describe how the extremely large configuration space of Fast Downward was restricted to a subspace that, although still very large, can be managed by state-of-the-art automated configuration procedures. FD-Autotune uses the well-known ParamILS configurator and was realised using the recently developed HAL experimentation environment.

• Malte Helmert, Gabriele Röger and Erez Karpas.
  Fast Downward Stone Soup: A Baseline for Building Planner Portfolios.
  In Proceedings of the ICAPS-2011 Workshop on Planning and Learning (PAL), pp. 28-35. 2011.
  (Show abstract) (Hide abstract) (PDF)
  

  Fast Downward Stone Soup is a sequential portfolio planner that uses various heuristics and search algorithms that have been implemented in the Fast Downward planning system.

  We present a simple general method for concocting "planner soups", sequential portfolios of planning algorithms, and describe the actual recipes used for Fast Downward Stone Soup in the sequential optimization and sequential satisficing tracks of IPC 2011.

  This paper is, first and foremost, a planner description. Fast Downward Stone Soup was entered into the sequential (non-learning) tracks of IPC 2011. Due to time constraints, we did not enter it into the learning competition at IPC 2011. However, we believe that the approach might still be of interest to the planning and learning community, as it represents a baseline against which other, more sophisticated portfolio learners can be usefully compared.

• Chris Fawcett, Malte Helmert, Holger Hoos, Erez Karpas, Gabriele Röger and Jendrik Seipp.
  FD-Autotune: Domain-Specific Configuration using Fast Downward.
  In Proceedings of the ICAPS-2011 Workshop on Planning and Learning (PAL), pp. 13-20. 2011.
  (Show abstract) (Hide abstract) (PDF)
  

  In this work, we present the FD-Autotune learning planning system, which is based on the idea of domain-specific configuration of the latest, highly parametric version of the Fast Downward Planning Framework by means of a generic automated algorithm configuration procedure. We describe how the extremely large configuration space of Fast Downward was restricted to a subspace that, although still very large, can be managed by a state-of-the-art automated configuration procedure. Additionally, we give preliminary results obtained from applying our approach to the nine domains of the IPC-2011 learning track, using the well-known ParamILS configurator and the recently developed HAL experimentation environment.

• Raz Nissim, Jörg Hoffmann and Malte Helmert.
  Computing Perfect Heuristics in Polynomial Time: On Bisimulation and Merge-and-Shrink Abstractions in Optimal Planning.
  In Proceedings of the ICAPS-2011 Workshop on Heuristics for Domain-independent Planning (HDIP), pp. 5-13. 2011.
  Superseded by the IJCAI 2011 paper by the same name.
  (Show abstract) (Hide abstract) (PDF)
  

  A* with admissible heuristics is a very successful approach to optimal planning. But how to derive such heuristics automatically? Merge-and-shrink abstraction (M&S) is a general approach to heuristic design whose key advantage is its capability to make very fine-grained choices in defining abstractions. However, little is known about how to actually make these choices. We address this via the well-known notion of bisimulation. When aggregating only bisimilar states, M&s yields a perfect heuristic. Alas, bisimulations are exponentially large even in trivial domains. We show how to apply label reduction -- not distinguishing between certain groups of operators -- without incurring any information loss, while potentially reducing bisimulation size exponentially. In several benchmark domains, the resulting algorithm computes perfect heuristics in polynomial time. Empirically, we show that approximating variants of this algorithm improve the state of the art in M&S heuristics. In particular, a hybrid of two such variants is competitive with the leading heuristic LM-cut.

• Jendrik Seipp and Malte Helmert.
  Fluent Merging for Classical Planning Problems.
  In Proceedings of the ICAPS-2011 Workshop on Knowledge Engineering for Planning and Scheduling (KEPS), pp. 47-53. 2011.
  Note: This version of the paper fixes two mistakes (in Def. 2 and in the text after Def. 3) that are present in the version of the paper that is linked from the workshop webpage..
  (Show abstract) (Hide abstract) (PDF)
  

  Fluent merging is a reformulation technique for classical planning problems that can be applied automatically or semi-automatically. The reformulation strives to transform a planning task into a representation that allows a planning algorithm to find solutions more efficiently or to find solutions of better quality. This work introduces different approaches for fluent merging and evaluates them within a state-of-the-art planning system.

• Fahiem Bacchus, Carmel Domshlak, Stefan Edelkamp and Malte Helmert (eds.).
  Proceedings of the 21st International Conference on Automated Planning and Scheduling (ICAPS 2011).
  AAAI Press, Menlo Park, California, USA 2011.
  

• Malte Helmert and Gabriele Röger.
  Relative-Order Abstractions for the Pancake Problem.
  In Helder Coelho, Rudi Studer and Michael Wooldridge (eds.), Proceedings of the 19th European Conference on Artificial Intelligence (ECAI 2010), pp. 745-750. IOS Press 2010.
  (Show abstract) (Hide abstract) (PDF)
  

  The pancake problem is a famous search problem where the objective is to sort a sequence of objects (pancakes) through a minimal number of prefix reversals (flips). The best approaches for the problem are based on heuristic search with abstraction (pattern database) heuristics. We present a new class of abstractions for the pancake problem called relative-order abstractions. Relative-order abstractions have three advantages over the object-location abstractions considered in previous work. First, they are size-independent, i.e., do not need to be tailored to a particular instance size of the pancake problem. Second, they are more compact in that they can represent a larger number of pancakes within abstractions of bounded size. Finally, they can exploit symmetries in the problem specification to allow multiple heuristic lookups, significantly improving search performance over a single lookup. Our experiments show that compared to object-location abstractions, our new techniques lead to an improvement of one order of magnitude in runtime and up to three orders of magnitude in the number of generated states.

• Blai Bonet and Malte Helmert.
  Strengthening Landmark Heuristics via Hitting Sets.
  In Helder Coelho, Rudi Studer and Michael Wooldridge (eds.), Proceedings of the 19th European Conference on Artificial Intelligence (ECAI 2010), pp. 329-334. IOS Press 2010.
  (Show abstract) (Hide abstract) (PDF) (technical report with proofs; PDF) (slides of Blai's ECAI 2010 presentation; PDF) (slides of Malte's SS 2010 group seminar presentation; PDF)
  

  The landmark cut heuristic is perhaps the strongest known polytime admissible approximation of the optimal delete relaxation heuristic h⁺. Equipped with this heuristic, a best-first search was able to optimally solve 40% more benchmark problems than the winners of the sequential optimization track of IPC 2008. We show that this heuristic can be understood as a simple relaxation of a hitting set problem, and that stronger heuristics can be obtained by considering stronger relaxations. Based on these findings, we propose a simple polytime method for obtaining heuristics stronger than landmark cut, and evaluate them over benchmark problems. We also show that hitting sets can be used to characterize h⁺ and thus provide a fresh and novel insight for better comprehension of the delete relaxation.

• Emil Keyder, Silvia Richter and Malte Helmert.
  Sound and Complete Landmarks for And/Or Graphs.
  In Helder Coelho, Rudi Studer and Michael Wooldridge (eds.), Proceedings of the 19th European Conference on Artificial Intelligence (ECAI 2010), pp. 335-340. IOS Press 2010.
  (Show abstract) (Hide abstract) (PDF)
  

  Landmarks for a planning problem are subgoals that are necessarily made true at some point in the execution of any plan. Since verifying that a fact is a landmark is PSPACE-complete, earlier approaches have focused on finding landmarks for the delete relaxation Π⁺. Furthermore, some of these approaches have approximated this set of landmarks, although it has been shown that the complete set of causal delete-relaxation landmarks can be identified in polynomial time by a simple procedure over the relaxed planning graph. Here, we give a declarative characterisation of this set of landmarks and show that the procedure computes the landmarks described by our characterisation. Building on this, we observe that the procedure can be applied to any delete-relaxation problem and take advantage of a recent compilation of the m-relaxation of a problem into a problem with no delete effects to extract landmarks that take into account delete effects in the original problem. We demonstrate that this approach finds strictly more causal landmarks than previous approaches and discuss the relationship between increased computational effort and experimental performance, using these landmarks in a recently proposed admissible landmark-counting heuristic.

• Malte Helmert.
  Lessons Learned from Benchmarking in the Automated Planning Community.
  In Proceedings of the ECAI 2010 Workshop on Benchmarking Intelligent (Multi-) Robot Systems. 2010.
  (PDF)
  

• Patrick Eyerich, Thomas Keller and Malte Helmert.
  High-Quality Policies for the Canadian Traveler's Problem.
  In Maria Fox and David Poole (eds.), Proceedings of the Twenty-Fourth AAAI Conference on Artificial Intelligence (AAAI 2010), pp. 51-58. AAAI Press 2010.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  We consider the stochastic variant of the Canadian Traveler's Problem, a path planning problem where adverse weather can cause some roads to be untraversable. The agent does not initially know which roads can be used. However, it knows a probability distribution for the weather, and it can observe the status of roads incident to its location. The objective is to find a policy with low expected travel cost.

  We introduce and compare several algorithms for the stochastic CTP. Unlike the optimistic approach most commonly considered in the literature, the new approaches we propose take uncertainty into account explicitly. We show that this property enables them to generate policies of much higher quality than the optimistic one, both theoretically and experimentally.

• Patrick Eyerich, Thomas Keller and Malte Helmert.
  High-Quality Policies for the Canadian Traveler's Problem (Extended Abstract).
  In Ariel Felner and Nathan Sturtevant (eds.), Proceedings of the Third Annual Symposium on Combinatorial Search (SoCS 2010), pp. 147-148. AAAI Press 2010.
  Extended abstract of the AAAI paper by the same name.
  (PDF)
  

• Malte Helmert.
  Landmark Heuristics for the Pancake Problem.
  In Ariel Felner and Nathan Sturtevant (eds.), Proceedings of the Third Annual Symposium on Combinatorial Search (SoCS 2010), pp. 109-110. AAAI Press 2010.
  (Show abstract) (Hide abstract) (PDF)
  

  We describe the gap heuristic for the pancake problem, which dramatically outperforms current abstraction-based heuristics for this problem. The gap heuristic belongs to a family of landmark heuristics that have recently been very successfully applied to planning problems.

• Malte Helmert and Hauke Lasinger.
  The Scanalyzer Domain: Greenhouse Logistics as a Planning Problem.
  In Ronen Brafman, Héctor Geffner, Jörg Hoffmann and Henry Kautz (eds.), Proceedings of the 20th International Conference on Automated Planning and Scheduling (ICAPS 2010), pp. 234-237. AAAI Press 2010.
  (Show abstract) (Hide abstract) (PDF)
  

  We introduce the Scanalyzer planning domain, a domain for classical planning which models the problem of automatic greenhouse logistic management.

  At its mathematical core, the Scanalyzer domain is a permutation problem with striking similarities to common search benchmarks such as Rubik's Cube or TopSpin. At the same time, it is also a real application domain, and efficient algorithms for the problem are of considerable practical interest.

  The Scanalyzer domain was used as a benchmark for sequential planners at the last International Planning Competition. The competition results show that domain-independent automated planners can find solutions of comparable quality to those generated by specialized algorithms developed by domain experts, while being considerably more flexible.

• Robert Mattmüller, Manuela Ortlieb, Malte Helmert and Pascal Bercher.
  Pattern Database Heuristics for Fully Observable Nondeterministic Planning.
  In Ronen Brafman, Héctor Geffner, Jörg Hoffmann and Henry Kautz (eds.), Proceedings of the 20th International Conference on Automated Planning and Scheduling (ICAPS 2010), pp. 105-112. AAAI Press 2010.
  (Show abstract) (Hide abstract) (PDF) (slides; PDF) (BIB)
  

  When planning in an uncertain environment, one is often interested in finding a contingent plan that prescribes appropriate actions for all possible states that may be encountered during the execution of the plan. We consider the problem of finding strong and strong cyclic plans for fully observable nondeterministic (FOND) planning problems. The algorithm we choose is LAO*, an informed explicit state search algorithm. We investigate the use of pattern database (PDB) heuristics to guide LAO* towards goal states. To obtain a fully domain-independent planning system, we use an automatic pattern selection procedure that performs local search in the space of pattern collections. The evaluation of our system on the FOND benchmarks of the Uncertainty Part of the International Planning Competition 2008 shows that our approach is competitive with symbolic regression search in terms of problem coverage and speed.

• Gabriele Röger and Malte Helmert.
  The More, the Merrier: Combining Heuristic Estimators for Satisficing Planning.
  In Ronen Brafman, Héctor Geffner, Jörg Hoffmann and Henry Kautz (eds.), Proceedings of the 20th International Conference on Automated Planning and Scheduling (ICAPS 2010), pp. 246-249. AAAI Press 2010.
  (Show abstract) (Hide abstract) (PDF) (technical report; PDF)
  

  We empirically examine several ways of exploiting the information of multiple heuristics in a satisficing best-first search algorithm, comparing their performance in terms of coverage, plan quality, speed, and search guidance. Our results indicate that using multiple heuristics for satisficing search is indeed useful. Among the combination methods we consider, the best results are obtained by the alternation method of the "Fast Diagonally Downward" planner.

• Patrick Eyerich, Thomas Keller and Malte Helmert.
  High-Quality Policies for the Canadian Traveler's Problem.
  In Proceedings of the ICAPS-2010 Workshop on Planning and Scheduling Under Uncertainty. 2010.
  Superseded by the AAAI 2010 paper by the same name.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  We consider the stochastic variant of the Canadian Traveler's Problem, a path planning problem where adverse weather can cause some roads to be untraversable. The agent does not initially know which roads can be used. However, it knows a probability distribution for the weather, and it can observe the status of roads incident to its location. The objective is to find a policy with low expected travel cost.

  We introduce and compare several algorithms for the stochastic CTP. Unlike the optimistic approach most commonly considered in the literature, the new approaches we propose take uncertainty into account explicitly. We show that this property enables them to generate policies of much higher quality than the optimistic one, both theoretically and experimentally.

• Dunbo Cai, Jörg Hoffmann and Malte Helmert.
  Enhancing the Context-Enhanced Additive Heuristic with Precedence Constraints.
  In Proceedings of the 19th International Conference on Automated Planning and Scheduling (ICAPS 2009), pp. 50-57. AAAI Press 2009.
  (Show abstract) (Hide abstract) (PDF)
  

  Recently, Helmert and Geffner proposed the context-enhanced additive heuristic, where fact costs are evaluated relative to context states that arise from achieving first a pivot condition of each operator. As Helmert and Geffner pointed out, the method can be generalized to consider contexts arising from arbitrary precedence constraints over operator conditions instead. Herein, we provide such a generalization. We extend Helmert and Geffner's equations, and discuss a number of design choices that arise. Drawing on previous work on goal orderings, we design a family of methods for automatically generating precedence constraints. We run large-scale experiments, showing that the technique can help significantly, depending on the choice of precedence constraints. We shed some light on this by profiling the behavior of all possible precedence constraints, using a sampling technique.

• Malte Helmert and Carmel Domshlak.
  Landmarks, Critical Paths and Abstractions: What's the Difference Anyway?
  In Proceedings of the 19th International Conference on Automated Planning and Scheduling (ICAPS 2009), pp. 162-169. AAAI Press 2009.
  (Show abstract) (Hide abstract) (PDF) (Dagstuhl abstract; PDF)
  

  Current heuristic estimators for classical domain-independent planning are usually based on one of four ideas: delete relaxations, critical paths, abstractions, and, most recently, landmarks. Previously, these different ideas for deriving heuristic functions were largely unconnected.

  We prove that admissible heuristics based on these ideas are in fact very closely related. Exploiting this relationship, we introduce a new admissible heuristic called the landmark cut heuristic, which compares favourably with the state of the art in terms of heuristic accuracy and overall performance.

• Silvia Richter and Malte Helmert.
  Preferred Operators and Deferred Evaluation in Satisficing Planning.
  In Proceedings of the 19th International Conference on Automated Planning and Scheduling (ICAPS 2009), pp. 273-280. AAAI Press 2009.
  (Show abstract) (Hide abstract) (PDF)
  

  Heuristic forward search is the dominant approach to satisficing planning to date. Most successful planning systems, however, go beyond plain heuristic search by employing various search-enhancement techniques. One example is the use of helpful actions or preferred operators, providing information which may complement heuristic values. A second example is deferred heuristic evaluation, a search variant which can reduce the number of costly node evaluations. Despite the wide-spread use of these search-enhancement techniques however, we note that few results have been published examining their usefulness. In particular, while various ways of using, and possibly combining, these techniques are conceivable, no work to date has studied the performance of such variations. In this paper, we address this gap by examining the use of preferred operators and deferred evaluation in a variety of settings within best-first search. In particular, our findings are consistent with and help explain the good performance of the winners of the satisficing tracks at IPC 2004 and 2008.

• Christoph Betz and Malte Helmert.
  Planning with h+ in Theory and Practice.
  In Proceedings of the 2nd Workshop on Heuristics for Domain-independent Planning at ICAPS 2009. 2009.
  (Show abstract) (Hide abstract) (PDF)
  

  Many heuristic estimators for classical planning are based on the so-called delete relaxation, which ignores negative effects of planning operators. Ideally, such heuristics would compute the actual goal distance in the delete relaxation, i.e., the cost of an optimal relaxed plan, denoted by h+. However, current delete relaxation heuristics only provide (often inadmissible) estimates to h+ because computing the correct value is an NP-hard problem.

  In this work, we consider the approach of planning with the actual h+ heuristic from a theoretical and computational perspective. In particular, we provide domain-dependent complexity results that classify some standard benchmark domains into ones where h+ can be computed efficiently and ones where computing h+ is NP-hard. Moreover, we study domain-dependent implementations of h+ which show that the h+ heuristic provides very informative heuristic estimates compared to other state-of-the-art heuristics.

• Gabriele Röger and Malte Helmert.
  Combining Heuristic Estimators for Satisficing Planning.
  In Proceedings of the 2nd Workshop on Heuristics for Domain-independent Planning at ICAPS 2009. 2009.
  (Show abstract) (Hide abstract) (PDF)
  

  The problem of effectively combining multiple heuristic estimators has been studied extensively in the context of optimal planning, but not in the context of satisficing planning. To narrow this gap, we empirically examine several ways of exploiting the information of multiple heuristics in a satisficing best-first search algorithm, comparing their performance in terms of coverage, plan quality and runtime. Our empirical results indicate that using multiple heuristics for satisficing search is indeed useful and that the best results are not obtained by the most obvious combination methods.

• Christoph Betz and Malte Helmert.
  Planning with h+ in Theory and Practice.
  In Bärbel Mertsching, Marcus Hund and Zaheer Aziz (eds.), Proceedings of the 32nd Annual German Conference on Artificial Intelligence (KI 2009), pp. 9-16. Springer-Verlag 2009.
  (Show abstract) (Hide abstract) (PDF)
  

  Many heuristic estimators for classical planning are based on the so-called delete relaxation, which ignores negative effects of planning operators. Ideally, such heuristics would compute the actual goal distance in the delete relaxation, i.e., the cost of an optimal relaxed plan, denoted by h+. However, current delete relaxation heuristics only provide (often inadmissible) estimates to h+ because computing the correct value is an NP-hard problem.

  In this work, we consider the approach of planning with the actual h+ heuristic from a theoretical and computational perspective. In particular, we provide domain-dependent complexity results that classify some standard benchmark domains into ones where h+ can be computed efficiently and ones where computing h+ is NP-hard. Moreover, we study domain-dependent implementations of h+ which show that the h+ heuristic provides very informative heuristic estimates compared to other state-of-the-art heuristics.

• Martin Wehrle and Malte Helmert.
  The Causal Graph Revisited for Directed Model Checking.
  In Jens Palsberg and Zhendong Su (eds.), Proceedings of the 16th International Static Analysis Symposium (SAS 2009), pp. 86-101. Springer-Verlag 2009.
  (Show abstract) (Hide abstract) (PDF) (Dagstuhl abstract; PDF)
  

  Directed model checking is a well-established technique to tackle the state explosion problem when the aim is to find error states in large systems. In this approach, the state space traversal is guided through a function that estimates the distance to nearest error states. States with lower estimates are preferably expanded during the search. Obviously, the challenge is to develop distance functions that are efficiently computable on the one hand and as informative as possible on the other hand. In this paper, we introduce the causal graph structure to the context of directed model checking. Based on causal graph analysis, we first adapt a distance estimation function from AI planning to directed model checking. Furthermore, we investigate an abstraction that is guaranteed to preserve error states. The experimental evaluation shows the practical potential of these techniques.

• Malte Helmert.
  Research Statement: Heuristic Search for Domain-Independent Planning.
  In 2nd International Symposium on Combinatorial Search (SoCS 2009). 2009.
  (PDF)
  

• Jörg Hoffmann, Piergiorgio Bertoli, Malte Helmert and Marco Pistore.
  Message-Based Web Service Composition, Integrity Constraints, and Planning under Uncertainty: A New Connection.
  Journal of Artificial Intelligence Research 35, pp. 49-117. 2009.
  (Show abstract) (Hide abstract) (PDF)
  

  Thanks to recent advances, AI Planning has become the underlying technique for several applications. Figuring prominently among these is automated Web Service Composition (WSC) at the "capability" level, where services are described in terms of preconditions and effects over ontological concepts. A key issue in addressing WSC as planning is that ontologies are not only formal vocabularies; they also axiomatize the possible relationships between concepts. Such axioms correspond to what has been termed "integrity constraints" in the actions and change literature, and applying a web service is essentially a belief update operation. The reasoning required for belief update is known to be harder than reasoning in the ontology itself. The support for belief update is severely limited in current planning tools.

  Our first contribution consists in identifying an interesting special case of WSC which is both significant and more tractable. The special case, which we term forward effects, is characterized by the fact that every ramification of a web service application involves at least one new constant generated as output by the web service. We show that, in this setting, the reasoning required for belief update simplifies to standard reasoning in the ontology itself. This relates to, and extends, current notions of "message-based" WSC, where the need for belief update is removed by a strong (often implicit or informal) assumption of "locality" of the individual messages. We clarify the computational properties of the forward effects case, and point out a strong relation to standard notions of planning under uncertainty, suggesting that effective tools for the latter can be successfully adapted to address the former.

  Furthermore, we identify a significant sub-case, named strictly forward effects, where an actual compilation into planning under uncertainty exists. This enables us to exploit off-the-shelf planning tools to solve message-based WSC in a general form that involves powerful ontologies, and requires reasoning about partial matches between concepts. We provide empirical evidence that this approach may be quite effective, using Conformant-FF as the underlying planner.

• Malte Helmert.
  Concise finite-domain representations for PDDL planning tasks.
  Artificial Intelligence 173, pp. 503-535. 2009.
  (Show abstract) (Hide abstract) (PDF)
  

  We introduce an efficient method for translating planning tasks specified in the standard PDDL formalism into a concise grounded representation that uses finite-domain state variables instead of the straight-forward propositional encoding.

  Translation is performed in four stages. Firstly, we transform the input task into an equivalent normal form expressed in a restricted fragment of PDDL. Secondly, we synthesize invariants of the planning task that identify groups of mutually exclusive propositions which can be represented by a single finite-domain variable. Thirdly, we perform an efficient relaxed reachability analysis using logic programming techniques to obtain a grounded representation of the input. Finally, we combine the results of the third and fourth stage to generate the final grounded finite-domain representation.

  The presented approach has originally been implemented as part of the Fast Downward planning system for the 4th International Planning Competition (IPC4). Since then, it has been used in a number of other contexts with considerable success, and the use of concise finite-domain representations has become a common feature of state-of-the-art planners.

• Gabriele Röger, Malte Helmert and Bernhard Nebel.
  On the Relative Expressiveness of ADL and Golog: The Last Piece in the Puzzle.
  In Proceedings of the Eleventh International Conference on Principles of Knowledge Representation and Reasoning (KR 2008), pp. 544-550. AAAI Press 2008.
  (Show abstract) (Hide abstract) (PDF)
  

  Integrating agent programming languages and efficient action planning is a promising approach because it combines the expressive power of languages such as Golog with the possibility of searching for plans efficiently. In order to integrate a Golog interpreter with a planner, one has to understand, however, which part of the expressiveness of Golog can be captured by the planning language. Using Nebel's compilation framework, we identify a maximal fragment of basic action theories, the formalism Golog is based on, that is expressively equivalent to the ADL subset of PDDL. As we will show, almost all features that permit to specify incomplete information in basic action theories cannot be compiled to ADL.

• Malte Helmert and Héctor Geffner.
  Unifying the Causal Graph and Additive Heuristics.
  In Proceedings of the 18th International Conference on Automated Planning and Scheduling (ICAPS 2008), pp. 140-147. AAAI Press 2008.
  (Show abstract) (Hide abstract) (PDF)
  

  Many current heuristics for domain-independent planning, such as Bonet and Geffner's additive heuristic and Hoffmann and Nebel's FF heuristic, are based on delete relaxations. They estimate the goal distance of a search state by approximating the solution cost in a relaxed task where negative consequences of operator applications are ignored. Helmert's causal graph heuristic, on the other hand, approximates goal distances by solving a hierarchy of "local" planning problems that only involve a single state variable and the variables it depends on directly.

  Superficially, the causal graph heuristic appears quite unrelated to heuristics based on delete relaxation. In this contribution, we show that the opposite is true. Using a novel, declarative formulation of the causal graph heuristic, we show that the causal graph heuristic is the additive heuristic plus context. Unlike the original heuristic, our formulation does not require the causal graph to be acyclic, and thus leads to a proper generalization of both the causal graph and additive heuristics. Empirical results show that the new heuristic is significantly better informed than both Helmert's original causal graph heuristic and the additive heuristic and outperforms them across a wide range of standard benchmarks.

• Malte Helmert, Patrik Haslum and Jörg Hoffmann.
  Explicit-State Abstraction: A New Method for Generating Heuristic Functions.
  In Proceedings of the 23rd AAAI Conference on Artificial Intelligence (AAAI 2008), pp. 1547-1550. AAAI Press 2008.
  (Show abstract) (Hide abstract) (PDF) (slides; PDF)
  

  Many AI problems can be recast as finding an optimal path in a discrete state space. An abstraction defines an admissible heuristic function as the distances in a smaller state space where arbitrary sets of states are "aggregated" into single states. A special case are pattern database (PDB) heuristics, which aggregate states iff they agree on the state variables inside the pattern. Explicit-state abstraction is more flexible, explicitly aggregating selected pairs of states in a process that interleaves composition of abstractions with abstraction of the composites. The increased flexibility gains expressive power: sometimes, the real cost function can be represented concisely as an explicit-state abstraction, but not as a PDB. Explicit-state abstraction has been applied to planning and model checking, with highly promising empirical results.

• Malte Helmert and Gabriele Röger.
  How Good is Almost Perfect?
  In Proceedings of the 23rd AAAI Conference on Artificial Intelligence (AAAI 2008), pp. 944-949. AAAI Press 2008.
  (Show abstract) (Hide abstract) (PDF) (slides; PDF)
  

  Heuristic search using algorithms such as A* and IDA* is the prevalent method for obtaining optimal sequential solutions for classical planning tasks. Theoretical analyses of these classical search algorithms, such as the well-known results of Pohl, Gaschnig and Pearl, suggest that such heuristic search algorithms can obtain better than exponential scaling behaviour, provided that the heuristics are accurate enough.

  Here, we show that for a number of common planning benchmark domains, including ones that admit optimal solution in polynomial time, general search algorithms such as A* must necessarily explore an exponential number of search nodes even under the optimistic assumption of almost perfect heuristic estimators, whose heuristic error is bounded by a small additive constant.

  Our results shed some light on the comparatively bad performance of optimal heuristic search approaches in "simple" domains such as GRIPPER. They suggest that in many domains, further improvements in run-time require changes to other parts of the planning algorithm than the heuristic estimator.

• Malte Helmert and Robert Mattmüller.
  Accuracy of Admissible Heuristic Functions in Selected Planning Domains.
  In Proceedings of the 23rd AAAI Conference on Artificial Intelligence (AAAI 2008), pp. 938-943. AAAI Press 2008.
  (Show abstract) (Hide abstract) (PDF) (slides; PDF) (BIB)
  

  The efficiency of optimal planning algorithms based on heuristic search crucially depends on the accuracy of the heuristic function used to guide the search. Often, we are interested in domain-independent heuristics for planning. In order to assess the limitations of domain-independent heuristic planning, it appears interesting to analyse the (in)accuracy of common domain-independent planning heuristics in the IPC benchmark domains. For a selection of these domains, we analytically investigate the accuracy of the h⁺ heuristic, the h^m family of heuristics, and certain (additive) pattern database heuristics, compared to the perfect heuristic h^*. Whereas h⁺ and additive pattern database heuristics usually return cost estimates proportional to the true cost, non-additive h^m and non-additive pattern-database heuristics can yield results underestimating the true cost by arbitrarily large factors.

• Silvia Richter, Malte Helmert and Matthias Westphal.
  Landmarks Revisited.
  In Proceedings of the 23rd AAAI Conference on Artificial Intelligence (AAAI 2008), pp. 975-982. AAAI Press 2008.
  Note: After publication, we found a bug in our implementation that affects the results in the columns "CG heuristic/local" and "blind heuristic/local" of Table 1. The version of the paper available for download here corrects these errors.
  (Show abstract) (Hide abstract) (PDF) (slides; PDF)
  

  Landmarks for propositional planning tasks are variable assignments that must occur at some point in every solution plan. We propose a novel approach for using landmarks in planning by deriving a pseudo-heuristic and combining it with other heuristics in a search framework. The incorporation of landmark information is shown to improve success rates and solution qualities of a heuristic planner. We furthermore show how additional landmarks and orderings can be found using the information present in multi-valued state variable representations of planning tasks. Compared to previously published approaches, our landmark extraction algorithm provides stronger guarantees of correctness for the generated landmark orderings, and our novel use of landmarks during search solves more planning tasks and delivers considerably better solutions.

• Malte Helmert.
  Understanding Planning Tasks: Domain Complexity and Heuristic Decomposition.
  Volume 4929 of Lecture Notes in Artificial Intelligence.
  Springer-Verlag, Heidelberg 2008.
  (Springer Online)
  

• Malte Helmert, Patrik Haslum and Jörg Hoffmann.
  Flexible Abstraction Heuristics for Optimal Sequential Planning.
  In Proceedings of the Seventeenth International Conference on Automated Planning and Scheduling (ICAPS 2007), pp. 176-183. AAAI Press 2007.
  (Show abstract) (Hide abstract) (PDF)
  

  We describe an approach to deriving consistent heuristics for automated planning, based on explicit search in abstract state spaces. The key to managing complexity is interleaving composition of abstractions over different sets of state variables with abstraction of the partial composites.

  The approach is very general and can be instantiated in many different ways by following different abstraction strategies. In particular, the technique subsumes planning with pattern databases as a special case. Moreover, with suitable abstraction strategies it is possible to derive perfect heuristics in a number of classical benchmark domains, thus allowing their optimal solution in polynomial time.

  To evaluate the practical usefulness of the approach, we perform empirical experiments with one particular abstraction strategy. Our results show that the approach is competitive with the state of the art.

• Malte Helmert and Gabriele Röger.
  How Good is Almost Perfect?
  In Proceedings of the ICAPS-2007 Workshop on Heuristics for Domain-independent Planning: Progress, Ideas, Limitations, Challenges. 2007.
  Superseded by the AAAI 2008 paper by the same name.
  (Show abstract) (Hide abstract) (PDF)
  

  Heuristic search using algorithms such as A* and IDA* is the prevalent method for obtaining optimal sequential solutions for classical planning tasks. Theoretical analyses of these classical search algorithms, such as the well-known results of Pohl, Gaschnig and Pearl, suggest that such heuristic search algorithms can obtain better than exponential scaling behaviour, provided that the heuristics are accurate enough.

  Here, we show that for a number of common planning benchmark domains, including ones that admit optimal solution in polynomial time, general search algorithms such as A* must necessarily explore an exponential number of search nodes even under the optimistic assumption of almost perfect heuristic estimators, whose heuristic error is bounded by a small additive constant.

  Our results shed some light on the comparatively bad performance of optimal heuristic search approaches in "simple" domains such as GRIPPER. They suggest that in many domains, further improvements in run-time require changes to other parts of the planning algorithm than the heuristic estimator.

• Malte Helmert and Robert Mattmüller.
  On the Accuracy of Admissible Heuristic Functions in Selected Planning Domains.
  In Proceedings of the ICAPS-2007 Workshop on Heuristics for Domain-independent Planning: Progress, Ideas, Limitations, Challenges. 2007.
  Superseded by the AAAI 2008 paper by the same name.
  (Show abstract) (Hide abstract) (PDF)
  

  The efficiency of optimal planning algorithms based on heuristic search crucially depends on the accuracy of the heuristic function used to guide the search. Often, we are interested in domain-independent heuristics for planning. In assessing the limitations of domain-independent heuristic planning, it appears interesting to analyse the (in)accuracy of common domain-independent planning heuristics in the IPC benchmark domains. For a selection of these domains, we analytically investigate the accuracy of the h⁺ heuristic, the h^k family of heuristics, and certain (additive) pattern database heuristics, compared to the optimal heuristic h^*. Whereas h⁺ and additive pattern database heuristics usually return cost estimates proportional to the true cost, non-additive h^k and non-additive pattern-database heuristics can yield results underestimating the true cost by arbitrarily large factors.

• Silvia Richter, Malte Helmert and Charles Gretton.
  A Stochastic Local Search Approach to Vertex Cover.
  In Proceedings of the 30th Annual German Conference on Artificial Intelligence (KI 2007), pp. 412-426. 2007.
  (Show abstract) (Hide abstract) (PDF)
  

  We introduce a novel stochastic local search algorithm for the vertex cover problem. Compared to current exhaustive search techniques, our algorithm achieves excellent performance on a suite of problems drawn from the field of biology. We also evaluate our performance on the commonly used DIMACS benchmarks for the related clique problem, finding that our approach is competitive with the current best stochastic local search algorithm for finding cliques. On three very large problem instances, our algorithm establishes new records in solution quality.

• Patrik Haslum, Adi Botea, Malte Helmert, Blai Bonet and Sven Koenig.
  Domain-Independent Construction of Pattern Database Heuristics for Cost-Optimal Planning.
  In Proceedings of the 22nd AAAI Conference on Artificial Intelligence (AAAI 2007), pp. 1007-1012. AAAI Press 2007.
  (Show abstract) (Hide abstract) (PDF)
  

  Heuristic search is a leading approach to domain-independent planning. For cost-optimal planning, however, existing admissible heuristics are generally too weak to effectively guide the search. Pattern database heuristics (PDBs), which are based on abstractions of the search space, are currently one of the most promising approaches to developing better admissible heuristics. The informedness of PDB heuristics depends crucially on the selection of appropriate abstractions (patterns). Although PDBs have been applied to many search problems, including planning, there are not many insights into how to select good patterns, even manually. What constitutes a good pattern depends on the problem domain, making the task even more difficult for domain-independent planning, where the process needs to be completely automatic and general. We present a novel way of constructing good patterns automatically from the specification of planning problem instances. We demonstrate that this allows a domain-independent planner to solve planning problems optimally in some very challenging domains, including a STRIPS formulation of the Sokoban puzzle.

• Malte Helmert, Robert Mattmüller and Sven Schewe.
  Selective Approaches for Solving Weak Games.
  In Proceedings of the Fourth International Symposium on Automated Technology for Verification and Analysis (ATVA 2006), pp. 200-214. Springer-Verlag 2006.
  (Show abstract) (Hide abstract) (PDF)
  

  Model-checking alternating-time properties has recently attracted much interest in the verification of distributed protocols. While checking the validity of a specification in alternating-time temporal logic (ATL) against an explicit model is cheap (linear in the size of the formula and the model), the problem becomes EXPTIME-hard when symbolic models are considered. Practical ATL model-checking therefore often consumes too much computation time to be tractable.

  In this paper, we describe a novel approach for ATL model-checking, which constructs an explicit weak model-checking game on-the-fly. This game is then evaluated using heuristic techniques inspired by efficient evaluation algorithms for and/or-trees.

  To show the feasibility of our approach, we compare its performance to the ATL model-checking system MOCHA on some practical examples. Using very limited heuristic guidance, we achieve a significant speedup on these benchmarks.

• Malte Helmert, Robert Mattmüller and Gabriele Röger.
  Approximation Properties of Planning Benchmarks.
  In Proceedings of the 17th European Conference on Artificial Intelligence (ECAI 2006), pp. 585-589. 2006.
  (Show abstract) (Hide abstract) (PDF)
  

  For many classical planning domains, the computational complexity of non-optimal and optimal planning is known. However, little is known about the area in between the two extremes of finding some plan and finding optimal plans. In this contribution, we provide a complete classification of the propositional domains from the first four International Planning Competitions with respect to the approximation classes PO, PTAS, APX, poly-APX, and NPO.

• Malte Helmert.
  New Complexity Results for Classical Planning Benchmarks.
  In Proceedings of the Sixteenth International Conference on Automated Planning and Scheduling (ICAPS 2006), pp. 52-61. AAAI Press 2006.
  (Show abstract) (Hide abstract) (PDF)
  

  The 3rd and 4th International Planning Competitions have enriched the set of benchmarks for classical propositional planning by a number of novel and interesting planning domains. Although they are commonly used for the purpose of evaluating planner performance, the planning domains themselves have not yet been studied in depth. In this contribution, we prove complexity results for the decision problems related to finding some plan, finding an optimal sequential plan, and finding an optimal parallel plan in these planning domains. Our results also provide some insights into the question why planning is hard for some of the domains under consideration.

• Sebastian Kupferschmid and Malte Helmert.
  A Skat Player Based on Monte Carlo Simulation.
  In H. Jaap van den Herik, Paolo Ciancarini and H. H. L. M. Donkers (eds.), Proceedings of the Fifth International Conference on Computer and Games (CG 2006), pp. 135-147. Springer-Verlag 2006.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  We apply Monte Carlo simulation and alpha-beta search to the card game of Skat, which is similar to Bridge, but different enough to require some new algorithmic ideas besides the techniques developed for Bridge. Our Skat-playing program integrates well-known techniques such as move ordering with two new search enhancements. Quasi-symmetry reduction generalizes symmetry reductions, popularized by Ginsberg's Partition Search algorithm, to search states which are "almost equivalent". Adversarial heuristics generalize ideas from single-agent search algorithms like A* to two-player games, leading to guaranteed lower and upper bounds for the score of a game position. Combining these techniques with state-of-the-art tree search algorithms, our program determines the game-theoretical value of a typical Skat hand (with perfect information) in 10 milliseconds.

• Malte Helmert.
  The Fast Downward Planning System.
  Journal of Artificial Intelligence Research 26, pp. 191-246. 2006.
  (Show abstract) (Hide abstract) (PDF)
  

  Fast Downward is a classical planning system based on heuristic search. It can deal with general deterministic planning problems encoded in the propositional fragment of PDDL2.2, including advanced features like ADL conditions and effects and derived predicates (axioms). Like other well-known planners such as HSP and FF, Fast Downward is a progression planner, searching the space of world states of a planning task in the forward direction. However, unlike other PDDL planning systems, Fast Downward does not use the propositional PDDL representation of a planning task directly. Instead, the input is first translated into an alternative representation called multi-valued planning tasks, which makes many of the implicit constraints of a propositional planning task explicit. Exploiting this alternative representation, Fast Downward uses hierarchical decompositions of planning tasks for computing its heuristic function, called the causal graph heuristic, which is very different from traditional HSP-like heuristics based on ignoring negative interactionse of operators.

  In this article, we give a full account of Fast Downward's approach to solving multi-valued planning tasks. We extend our earlier discussion of the causal graph heuristic to tasks involving axioms and conditional effects and present some novel techniques for search control that are used within Fast Downward's best-first search algorithm: preferred operators transfer the idea of helpful actions from local search to global best-first search, deferred evaluation of heuristic functions mitigates the negative effect of large branching factors on search performance, and multi-heuristic best-first search combines several heuristic evaluation functions within a single search algorithm in an orthogonal way. We also describe efficient data structures for fast state expansion (successor generators and axiom evaluators) and present a new non-heuristic search algorithm called focused iterative-broadening search, which utilizes the information encoded in causal graphs in a novel way.

  Fast Downward has proven remarkably successful: It won the "classical" (i.e., propositional, non-optimising) track of the 4th International Planning Competition at ICAPS 2004, following in the footsteps of planners such as FF and LPG. Our experiments show that it also performs very well on the benchmarks of the earlier planning competitions and provide some insights about the usefulness of the new search enhancements.

• Malte Helmert.
  A Planning Heuristic Based on Causal Graph Analysis.
  In Proceedings of the Fourteenth International Conference on Automated Planning and Scheduling (ICAPS 2004), pp. 161-170. AAAI Press 2004.
  (Show abstract) (Hide abstract) (PDF)
  

  In recent years, heuristic search methods for classical planning have achieved remarkable results. Their most successful representative, the FF algorithm, performs well over a wide spectrum of planning domains and still sets the state of the art for STRIPS planning. However, there are some planning domains in which algorithms like FF and HSP perform poorly because their relaxation method of ignoring the "delete lists" of STRIPS operators loses too much vital information.

  Planning domains which have many dead ends in the search space are especially problematic in this regard. In some domains, dead ends are readily found by the human observer yet remain undetected by all propositional planning systems we are aware of. We believe that this is partly because the STRIPS representation obscures the important causal structure of the domain, which is evident to humans.

  In this paper, we propose translating STRIPS problems to a planning formalism with multi-valued state variables in order to expose this underlying causal structure. Moreover, we show how this structure can be exploited by an algorithm for detecting dead ends in the search space and by a planning heuristic based on hierarchical problem decomposition.

  Our experiments show excellent overall performance on the benchmarks from the international planning competitions.

• Malte Helmert.
  Complexity results for standard benchmark domains in planning.
  Artificial Intelligence 143 (2), pp. 219-262. 2003.
  (Show abstract) (Hide abstract) (PDF)
  

  The efficiency of AI planning systems is usually evaluated empirically. For the validity of conclusions drawn from such empirical data, the problem set used for evaluation is of critical importance. In planning, this problem set usually, or at least often, consists of tasks from the various planning domains used in the first two international planning competitions, hosted at the 1998 and 2000 AIPS conferences. It is thus surprising that comparatively little is known about the properties of these benchmark domains, with the exception of BLOCKSWORLD, which has been studied extensively by several research groups.

  In this contribution, we try to remedy this fact by providing a map of the computational complexity of non-optimal and optimal planning for the set of domains used in the competitions. We identify a common transportation theme shared by the majority of the benchmarks and use this observation to define and analyze a general transportation problem that generalizes planning in several classical domains such as LOGISTICS, MYSTERY and GRIPPER. We then apply the results of that analysis to the actual transportation domains from the competitions. We next examine the remaining benchmarks, which do not exhibit a strong transportation feature, namely SCHEDULE and FREECELL.

  Relating the results of our analysis to empirical work on the behavior of the recently very successful FF planning system, we observe that our theoretical results coincide well with data obtained from empirical investigations.

• Malte Helmert.
  Decidability and Undecidability Results for Planning with Numerical State Variables.
  In M. Ghallab, J. Hertzberg and P. Traverso (eds.), Proceedings of the Sixth International Conference on Artificial Intelligence Planning and Scheduling (AIPS 2002), pp. 303-312. AAAI Press 2002.
  (Show abstract) (Hide abstract) (PDF)
  

  These days, propositional planning can be considered a quite well-understood problem. Good algorithms are known that will solve a wealth of very different and sometimes challenging planning tasks, and theoretical computational properties of both general STRIPS-style planning and the best-known benchmark problems have been established.

  However, propositional planning has a major drawback: The formalism is too weak to allow for the easy encoding of many genuinely interesting planning problems, specifically those involving numbers. A recent effort to enhance the PDDL planning language to cope with (among other additions) numerical state variables, to be used at the third international planning competition, has increased interest in these issues.

  In this contribution, we analyze "STRIPS with numbers" from a theoretical point of view. Specifically, we show that the introduction of numerical state variables makes the planning problem undecidable in the general case and many restrictions thereof and identify special cases for which we can provide decidability results.

• Stefan Edelkamp and Malte Helmert.
  The Model Checking Integrated Planning System (MIPS).
  AI Magazine 22 (3), pp. 67-71. 2001.
  (Show abstract) (Hide abstract) (PDF)
  

  MIPS was the first general planning system based on model checking methods. It can handle the STRIPS subset of the PDDL language and some additional features from ADL, namely negative preconditions and (universal) conditional effects. At the AIPS 2000 conference, MIPS was one of five planning systems to be awarded for "Distinguished Performance" in the fully automated track.

  This short paper gives a brief introduction to BDDs and explains the basic planning algorithm employed by MIPS, using a simple logistics problem as an example.

• Malte Helmert.
  On the Complexity of Planning in Transportation Domains.
  In A. Cesta and D. Borrajo (eds.), Proceedings of the 6th European Conference on Planning (ECP 2001), pp. 349-360. 2001.
  (Show abstract) (Hide abstract) (PDF)
  

  The efficiency of AI planning systems is usually evaluated empirically. Some benchmark problems are of particular importance in this context, especially the ones used in the competitions of the 1998 and 2000 AIPS conferences. Many of these benchmarks share a common theme of transporting portables, making use of mobiles traversing a map of locations and roads.

  In this contribution, we embed these benchmarks into a well-structured hierarchy of transportation problems and study the computational complexity of optimal and non-optimal planning in this family of domains. We identify the key domain features that make transportation tasks hard and try to shed some light on the recent success of planning systems based on heuristic local search, as observed in the AIPS 2000 competition.

Andreas Hertle

• Andreas Hertle and Bernhard Nebel.
  Efficient Auction Based Coordination for Distributed Multi-Agent Planning in Temporal Domains Using Resource Abstraction.
  In Proceedings of the 41st German Conference on Artificial Intelligence (KI 2018). 2018.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  Recent advances in mobile robotics and AI promise to revolutionize industrial production. As autonomous robots are able to solve more complex tasks, the difficulty of integrating various robot skills and coordinating groups of robots increases dramatically. Domain independent planning promises a possible solution. For single robot systems a number of successful demonstrations can be found in scientific literature. However our experiences at the RoboCup Logistics League in 2017 highlighted a severe lack in plan quality when coordinating multiple robots.

  In this work we demonstrate how out of the box temporal planning systems can be employed to increase plan quality for temporal multi-robot tasks. An abstract plan is generated first and sub-tasks in the plan are auctioned off to robots, which in turn employ planning to solve these tasks and compute bids. We evaluate our approach on two planning domains and find significant improvements in solution coverage and plan quality.

• Andreas Hertle and Bernhard Nebel.
  Identifying Good Poses When Doing Your Household Chores: Creation and Exploitation of Inverse Surface Reachability Maps.
  In Proceedings of the 2017 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS 2017). 2017.
  (Show abstract) (Hide abstract) (PDF)
  

  In current approaches to combined task and motion planning, usually symbolic planning and sampling based motion-planning are integrated. One problem is here to come up with good samples. We address the problem of identifying useful poses for a robot close to working surfaces such as tables or shelves. Our approach is based on reachability inversion which answers the question: where should the robot be located in order to reach a certain object? We extend the concept from point-based objects to flat polygonal surfaces in order to enable the robot to have a a good grasping position for many objects. Our approach allows to quickly sample multiple distinct poses for the robot from an prior computed distribution. Further we show how sampling from an inverse reachability distribution can be integrated into a CTAMP system.

• Christian Dornhege, Alexander Kleiner, Andreas Hertle and Andreas Kolling.
  Multirobot Coverage Search in Three Dimensions.
  Journal of Field Robotics 33 (4), pp. 537-558. 2016.
  (Show abstract) (Hide abstract) (PDF)
  

  Searching for objects and observing parts of a known environment efficiently is a fundamental problem in many real-world robotic applications, e.g., household robots searching for objects, inspection robots searching for leaking pipelines, and rescue robots searching for survivors after a disaster. We consider the problem of identifying and planning sequences of sensor locations from which robot sensors can observe and cover complex three-dimensional (3D) environments while traveling only short distances. Our approach is based on sampling and ranking a large number of sensor locations for a 3D environment represented by an OctoMap. The visible area from these sensor locations induces a minimal partition of the 3D environment that we exploit for planning sequences of sensor locations with short travel times efficiently. We present multiple planning algorithms designed for single robots and for multirobot teams. These algorithms include variants that are greedy, optimal, or based on decomposing the planning problem into a set cover and traveling salesman problem. We evaluated and compared these algorithms empirically in simulation and real-world robot experiments with up to four robots. Our results demonstrate that, despite the intractability of the overall problem, computing and executing effective solutions for multirobot coverage search in real 3D environments is feasible and ready for real-world applications.

• Armin Hornung, Sebastian Boettcher, Christian Dornhege, Andreas Hertle, Jonas Schlagenhauf and Maren Bennewitz.
  Mobile Manipulation in Cluttered Environments with Humanoids: Itegrated Perception, Task Planning, and Action Execution.
  In Proceedings of the IEEE-RAS International Conference on Humanoid Robots (HUMANOIDS). 2014.
  (Show abstract) (Hide abstract) (PDF)
  

  To autonomously carry out complex mobile manipulation tasks, a robot control system has to integrate several components for perception, world modeling, action planning and replanning, navigation, and manipulation. In this paper, we present a modular framework that is based on the Temporal Fast Downward Planner and supports external modules to control the robot. This allows to tightly integrate individual sub-systems with the high-level symbolic planner and enables a humanoid robot to solve challenging mobile manipulation tasks. In the work presented here, we address mobile manipulation with humanoids in cluttered environments, particularly the task of collecting objects and delivering them to designated places in a home-like environment while clearing obstacles out of the way. We implemented our system for a Nao humanoid tidying up a room, i.e., the robot has to collect items scattered on the floor, move obstacles out of its way, and deliver the objects to designated target locations. Despite the limited sensing and motion capabilities of the low-cost platform, the experiments show that our approach results in reliable task execution by applying monitoring actions to verify object and robot states.

• Andreas Hertle, Christian Dornhege, Thomas Keller, Robert Mattmüller, Manuela Ortlieb and Bernhard Nebel.
  An Experimental Comparison of Classical, FOND and Probabilistic Planning.
  In Proceedings of the 37th German Conference on Artificial Intelligence (KI 2014), pp. 297-308. Springer 2014.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  Domain-independent planning in general is broadly applicable to a wide range of tasks. Many formalisms exist that allow to describe different aspects of realistic problems. Which one is best is not clear as usually more expressiveness comes with a cost in planning time. Under the assumption that hard guarantees are not required, users are faced with a decision between multiple approaches. In this work we study the effect of nondeterministic action outcomes. As a generic model we use a probabilistic description in the form of Markov Decision Processes (MDPs). We define abstracting translations into a classical planning formalism and fully observable nondeterministic (FOND) planning. Our goal is to give insight into how state-of-the-art systems perform on different MDP planning domains. We evaluate those MDPs by running state-of-the-art planning systems on the abstracted formalisms.

• Christian Dornhege, Andreas Hertle and Bernhard Nebel.
  Lazy Evaluation and Subsumption Caching for Search-Based Integrated Task and Motion Planning.
  In Proceedings of the IROS workshop on AI-based robotics. 2013.
  (Show abstract) (Hide abstract) (PDF)
  

  State of the art classical planning systems can efficiently solve large symbolic problem instances. Applying classical planning techniques to robotics is possible by in- tegrating geometric reasoning in the planning process. The problems that are solvable in this way are significantly smaller than purely logical formulations as many costly geometric calculations are requested by a planner. Therefore we aim to avoid those calculations while preserving correctness. We address this problem with efficient caching techniques. Subsumption caching avoids costly computations by caching geometric queries and beyond answering the same queries also considers less or more constrained ones. Additionally, we describe a lazy evaluation technique that pushes applicability checks for successor states performing geometric queries to a later point. As we are interested in the performance of our planner not as a standalone component, but as part of an intelligent robotic system, we evaluate those techniques embedded in an integrated system during real-world mobile manipulation experiments.

• Tim Niemueller, Nichola Abdo, Andreas Hertle, Gerhard Lakemeyer, Wolfram Burgard and Bernhard Nebel.
  Towards Deliberative Active Perception using Persistent Memory.
  In Proceedings of the IROS workshop on AI-based robotics. 2013.
  (Show abstract) (Hide abstract) (PDF)
  

  Task coordination for autonomous mobile service robots typically involves a substantial amount of background knowledge and explicit action sequences to acquire the relevant information nowadays. We strive for a system which, given a task, is capable of reasoning about task-relevant knowledge to automatically determine whether that knowledge is sufficient. If missing or uncertain, the robot shall decide autonomously on the actions to gain or improve that knowledge. In this paper we present our baseline system implementing the foundations for these capabilities. The robot has to analyze a tabletop scene and increase its object type confidence. It plans motions to observe the scene from multiple perspectives, combines the acquired data, and performs a recognition step on the merged input.

• Andreas Hertle and Christian Dornhege.
  Efficient Extensible Path Planning on 3D Terrain Using Behavior Modules.
  In Proceedings of the European Conference on Mobile Robotics (ECMR). 2013.
  (PDF) (BIB)
  

• Bernhard Nebel, Christian Dornhege and Andreas Hertle.
  How Much Does a Household Robot Need To Know In Order To Tidy Up Your Home?
  In AAAI Workshop on Intelligent Robotic Systems. AAAI Press 2013.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  Although planning for the tasks a household robot has to perform appears to be easy, there exists the problem that the robot is usually uncertain about the state of the household when starting to plan. For example, when getting the order of tidying up the kitchen, the robot does not know what objects it will have to put away and whether there are actually any objects that need to be put away. Furthermore, while sensing operations can provide more information about the environment, things can go wrong when executing an action. In this paper, we try to identify conditions under which classical planning can be used in a replanning loop in order to solve the planning problem in nondeterminis- tic partially observable open domains. In particular, we will define completeness and soundness of replanning with respect to nondeterministic planning and we will identify a PSPACE-checkable condition that guarantees soundness.

• Christian Dornhege and Andreas Hertle.
  Integrated Symbolic Planning in the Tidyup-Robot Project.
  In AAAI Spring Symposium - Designing Intelligent Robots: Reintegrating AI II. AAAI Press 2013.
  (PDF) (BIB)
  

• Stefan Kohlbrecher, Karen Petersen, Gerald Steinbauer, Johannes Maurer, Peter Lepej, Suzana Uran, Rodrigo Ventura, Christian Dornhege, Andreas Hertle, Raymond Sheh and Johannes Pellenz.
  Community-Driven Development of Standard Software Modules for Search and Rescue Robots.
  In Safety, Security and Rescue Robotics (SSRR). 2012.
  

• Andreas Hertle, Christian Dornhege, Thomas Keller and Bernhard Nebel.
  Planning with Semantic Attachments: An Object-Oriented View.
  In Proceedings of the European Conference on Artificial Intelligence (ECAI). 2012.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  In recent years, domain-independent planning has been applied to a rising number of real-world applications. Usually, the description language of choice is PDDL. However, PDDL is not suited to model all challenges imposed by real-world applications. Dornhege et al. proposed semantic attachments to allow the computation of Boolean fluents by external processes called modules during planning. To acquire state information from the planning system a module developer must perform manual requests through a callback interface which is both inefficient and error-prone. In this paper, we present the Object-oriented Planning Language OPL, which incorporates the structure and advantages of modern object-oriented programming languages. We demonstrate how a domain-specific module interface that allows to directly access the planner state using object member functions is automatically gen- erated from an OPL planning task. The generated domain-specific interface allows for a safe and less error-prone implementation of modules. We show experimentally that this interface is more efficient than the PDDL-based module interface of TFD/M.

Jörg Hoffmann

• Sylvie Thiebaux, Jörg Hoffmann and Bernhard Nebel.
  In Defense of PDDL Axioms.
  Artificial Intelligence 168 (1-2), pp. 38-69. 2005.
  (Show abstract) (Hide abstract) (PDF)
  

  There is controversy as to whether explicit support for PDDL-like axioms and derived predicates is needed for planners to handle real-world domains effectively. Many researchers have deplored the lack of precise semantics for such axioms, while others have argued that it might be best to compile them away. We propose an adequate semantics for PDDL axioms and show that they are an essential feature by proving that it is impossible to compile them away if we restrict the growth of plans and domain descriptions to be polynomial. These results suggest that adding a reasonable implementation to handle axioms inside the planner is beneficial for the performance. Our experiments confirm this suggestion.

• Jörg Hoffmann and Sebastian Kupferschmid.
  A Covering Problem for Hypercubes.
  In Leslie Pack Kaelbling and Alessandro Saffiotti (eds.), Poster Proceedings of the 19th International Joint Conference on Artificial Intelligence (IJCAI 2005), pp. 1523-1524. 2005.
  (Show abstract) (Hide abstract) (PDF) (PS.GZ) (BIB)
  

  We introduce a new NP-complete problem asking if a "query" hypercube is (not) covered by a set of other "evidence" hypercubes. This comes down to a form of constraint reasoning asking for the satisfiability of a CNF formula where the logical atoms are inequalities over single variables, with possibly infinite variable domains. We empirically investigate the location of the phase transition regions in two random distributions of problem instances. We introduce a solution method that iteratively constructs a representation of the non-covered part of the query cube. In particular, the method is not based on backtracking. Our experiments show that the method is, in a significant range of instances, superior to the backtracking method that results from translation to SAT, and application of a state-of-the-art DP-based SAT solver.

• Sebastian Trüg, Jörg Hoffmann and Bernhard Nebel.
  Applying Automatic Planning Techniques to Airport Ground-Traffic Control: A Feasibility Study.
  In S. Biundo, T. Frühwirth and G. Palm (eds.), KI 2004: Advances in Artificial Intelligence. Proceedings of the 27th Annual German Conference on Artificial Intelligence, pp. 183-197. Springer-Verlag 2004.
  (Show abstract) (Hide abstract) (PDF)
  

  Planning techniques have matured as demonstrated by the performance of automatic planning systems at recent international competitions. Nowadays it seems feasible to apply planning systems to real-world problems. In order to get an idea of what the performance difference between special-purpose techniques and automated planning techniques is, we applied these techniques to the airport traffic control problem and compared it with a special purpose tool. In addition to a perfomance assessment, this exercise also resulted in a domain model of the airport traffic control domain. which was used as a benchmark in the 4th International Planning Competition.

• Jörg Hoffmann, Julie Porteous and Laura Sebastia.
  Ordered Landmarks in Planning.
  Journal of Artificial Intelligence Research 22, pp. 215-278. 2004.
  (PS.GZ)
  

• Ronen Brafman and Jörg Hoffmann.
  Conformant Planning via Heuristic Forward Search: A New Approach.
  In Proceedings of the Fourteenth International Conference on Automated Planning and Scheduling (ICAPS 2004), pp. 355-364. AAAI Press 2004.
  (PS.GZ)
  

• Bernd Becker, Markus Behle, Fritz Eisenbrand, Martin Fränzle, Marc Herbstritt, Christian Herde, Jörg Hoffmann, Daniel Kröning, Bernhard Nebel, Ilia Polian and Ralf Wimmer.
  Bounded Model Checking and Inductive Verification of Hybrid Discrete-continuous Systems.
  In Proceedings GI/ITG/GMM-Workshop Methoden und Beschreibungssprachen zur Modellierung und Verifikation von Schaltungen und Systemen, pp. 65-75. Kaiserslautern 2004.
  (Show abstract) (Hide abstract) (PDF)
  

  We present a concept to signicantly advance the state of the art for bounded model checking (BMC) and inductive verication (IV) of hybrid discrete-continuous systems. Our approach combines the expertise of partners coming from dierent domains, like hybrid systems modeling and digital circuit verication, bounded planning and heuristic search, combinatorial optimization and integer programming. After sketching the overall verication ow we present rst results indicating that the combination and tight integration of dierent verication engines is a rst step to pave the way to fully automated BMC and IV of medium to large-scale networks of hybrid automata.

• Jörg Hoffmann.
  Utilizing Problem Structure in Planning: A Local Search Approach.
  Volume 2854 of Lecture Notes in Artificial Intelligence.
  Springer-Verlag, Berlin, Heidelberg, New York 2003.
  (Springer Online) (extended abstract; PS.GZ)
  

• Ronen Brafman and Jörg Hoffmann.
  Conformant Planning via Heuristic Forward Search.
  In Proceedings of the Workshop on Planning under Uncertainty and Incomplete Information at ICAPS'03. Trento, Italy 2003.
  (PS.GZ)
  

• Stefan Edelkamp and Jörg Hoffmann.
  Quo Vadis, IPC-4? - Proposals for the Classical Part of the 4th International Planning Competition.
  In Proceedings of the Workshop on the Competition at ICAPS'03. Trento, Italy 2003.
  (PS.GZ)
  

• Jörg Hoffmann.
  The Metric-FF Planning System: Translating "Ignoring Delete Lists" to Numeric State Variables.
  Journal of Artificial Intelligence Research Special issue on the 3rd International Planning Competition. 2003.
  (PS.GZ)
  

• Jörg Hoffmann and Hector Geffner.
  Branching Matters: Alternative Branching in Graphplan.
  In Proceedings of the Thirteenth International Conference on Automated Planning and Scheduling. Trento, Italy 2003.
  (PS.GZ)
  

• Sylvie Thiebaux, Jörg Hoffmann and Bernhard Nebel.
  In Defense of PDDL Axioms.
  In Proceedings of the 18th International Joint Conference on Artificial Intelligence. Acapulco, Mexico 2003.
  (Show abstract) (Hide abstract) (PDF)
  

  There is controversy as to whether explicit support for PDDL-like axioms and derived predicates is needed for planners to handle real-world domains effectively. Many researchers have deplored the lack of precise semantics for such axioms, while others have argued that they are a non-essential feature which is best compiled away. We propose an adequate semantics for PDDL axioms and show that they are an essential feature by proving that it is impossible to compile them away if we restrict the growth of plans and domain descriptions to be polynomial. These results suggest that adding a reasonable implementation to handle axioms inside the planner is beneficial for the performance. Our experiments confirm this suggestion.

• Jörg Hoffmann.
  Local Search Topology in Planning Benchmarks: A Theoretical Analysis.
  In M. Ghallab, J. Hertzberg and P. Traverso (eds.), Proceedings of the Sixth International Conference on Artificial Intelligence Planning and Scheduling (AIPS 2002). AAAI Press 2002.
  (PS.GZ) (PDF)
  

• Jörg Hoffmann.
  Extending FF to Numerical State Variables.
  In Proceedings of the 15th European Conference on Artificial Intelligence. Lyon, France 2002.
  (PS.GZ)
  

• Jörg Hoffmann and Bernhard Nebel.
  RIFO revisited: Detecting Relaxed Irrelevance.
  In A. Cesta and D. Borrajo (eds.), Proceedings of the 6th European Conference on Planning (ECP 2001). 2001.
  (Show abstract) (Hide abstract) (PDF)
  

  RIFO, as has been proposed by Nebel et al., is a method that can automatically detect irrelevant information in planning tasks. The idea is to remove such irrelevant information as a pre­process to planning. While RIFO has been shown to be useful in a number of domains, its main disadvantage is that it is not completeness preserving. Furthermore, the pre­process often takes more running time than nowadays state­of­the­art planners, like FF, need for solving the entire planning task. We introduce the notion of relaxed irrelevance, concerning actions which are never needed within the relaxation that heuristic planners like FF and HSP use for computing their heuristic values. The idea is to speed up the heuris­ tic functions by reducing the action sets considered within the relaxation. Starting from a sufficient condition for relaxed irrelevance, we introduce two preprocessing methods for filtering action sets. The first preprocessing method is proven to be completeness­preserving, and is empirically shown to terminate fast on most of our testing examples. The second method is fast on all our testing examples, and is empirically safe. Both methods have drastic pruning impacts in some domains, speeding up FF's heuristic function, and in effect the planning process.

• Julie Porteous, Laura Sebastia and Jörg Hoffmann.
  On the Extraction, Ordering, and Usage of Landmarks in Planning.
  In A. Cesta and D. Borrajo (eds.), Proceedings of the 6th European Conference on Planning (ECP 2001). 2001.
  (PS.GZ) (PDF)
  

• Jörg Hoffmann.
  FF: The Fast-Forward Planning System.
  AI Magazine 22 (3), pp. 57-62. 2001.
  (PS.GZ) (PDF)
  

• Jörg Hoffmann and Bernhard Nebel.
  The FF Planning System: Fast Plan Generation Through Heuristic Search.
  Journal of Artificial Intelligence Research 14, pp. 253-302. 2001.
  (Show abstract) (Hide abstract) (PDF)
  

  We describe and evaluate the algorithmic techniques that are used in the FF planning system. Like the HSP system, FF relies on forward state space search, using a heuristic that estimates goal distances by ignoring delete lists. Unlike HSP's heuristic, our method does not assume facts to be independent. We introduce a novel search strategy that combines hill-climbing with systematic search, and we show how other powerful heuristic information can be extracted and used to prune the search space. FF was the most successful automatic planner at the recent AIPS-2000 planning competition. We review the results of the competition, give data for other benchmark domains, and investigate the reasons for the runtime performance of FF compared to HSP.

• Jörg Hoffmann.
  Local Search Topology in Planning Benchmarks: An Empirical Analysis.
  In Proceedings of the 17th International Joint Conference on Artificial Intelligence. Seattle, Washington, USA 2001.
  (PS.GZ) (PDF)
  

• Jörg Hoffmann and Bernhard Nebel.
  What makes the difference between HSP and FF?
  In IJCAI Workshop on Empirical AI. Seattle 2001.
  (Show abstract) (Hide abstract) (PDF)
  

  The HSP and FF systems are state-of-the-art domain independent planners. FF can historically be seen as a successor of HSP. It is based on the same ideas like HSP, but differs from its predecessor in a number of details. FF outperforms HSP in many planning domains. We have carried out a large scale experiment where we ran all configurations of FF's new techniques on a sizeable set of planning tasks. We describe the experimental design, and present our findings. The results give a clear picture of what the most important reasons are for FF's performance advantage over HSP.

• Jörg Hoffmann and Bernhard Nebel.
  Towards Thorough Empirical Methods for AI Planning.
  In IJCAI Workshop on Empirical AI. Seattle 2001.
  (Show abstract) (Hide abstract) (PDF)
  

  Empirical investigations in the area of AI planning have been focused on a comparatively small set of benchmark tasks. Trying to design larger scale experiments for well-founded empirical reasoning in that area, one encounters a number of severe problems. While some of these problems are inherent to the field, others have plainly been ignored. In our own work, we have made some first steps towards addressing these problems.

• Jussi Rintanen and Jörg Hoffmann.
  An overview of recent algorithms for AI planning.
  Künstliche Intelligenz Heft 2/01, pp. 5-11. 2001.
  (PS.GZ) (PDF)
  

• Jörg Hoffmann.
  A Heuristic for Domain Independent Planning and its Use in an Enforced Hill-climbing Algorithm.
  In Proceedings of the 14th Workshop on New Results in Planning, Scheduling and Design (PuK 2000) at ECAI 2000, pp. 62-67. Berlin, Germany 2000.
  (PS.GZ)
  

• Jana Koehler and Jörg Hoffmann.
  On the Instantation of ADL Operators Involving Arbitrary First-Order Formulas.
  In Proceedings of the 14th Workshop on New Results in Planning, Scheduling and Design (PuK 2000) at ECAI 2000, pp. 74-82. Berlin, Germany 2000.
  (PS.GZ)
  

• Jörg Hoffmann.
  A Heuristic for Domain Independent Planning and its Use in an Enforced Hill-climbing Algorithm.
  In Proceedings of the 12th International Symposium on Methodologies for Intelligent Systems. Charlotte, North Carolina, USA 2000.
  (PS.GZ)
  

• Jana Koehler and Jörg Hoffmann.
  On Reasonable and Forced Goal Orderings and their Use in an Agenda-Driven Planning Algorithm.
  Journal of Artificial Intelligence Research 12, pp. 338-386. 2000.
  (PS.GZ)
  

• Jana Koehler and Jörg Hoffmann.
  Planning with Goal Agendas.
  In Proceedings des 13. Workshops Planen und Konfigurieren (PuK 1999) auf der 10. Tagung Expertensysteme (XPS-99). Würzburg, Germany 1999.
  (PS.GZ)
  

• Jörg Hoffmann and Jana Koehler.
  A new Method to Query and Index Sets.
  In Proceedings of the 16th International Joint Conference on Artificial Intelligence (IJCAI 1999). Stockholm, Sweden 1999.
  (PS.GZ) (extended technical report; PS.GZ)
  

• Jana Koehler, Bernhard Nebel, Jörg Hoffmann and Yannis Dimopoulos.
  Extending Planning Graphs to an ADL Subset.
  In Proc. European Conference on Planning 1997 (ECP-97), pp. 273-285. Springer-Verlag 1997.
  (Show abstract) (Hide abstract) (PDF)
  

  We describe an extension of GRAPHPLAN to a subset of ADL that allows conditional and universally quantified effects in operators in such a way that almost all interesting properties of the original Graphplan algorithm are preserved.

Julien Hué

• Nicolas Riesterer, Christian Becker-Asano, Julien Hué, Christian Dornhege and Bernhard Nebel.
  The Hybrid Agent MARCO.
  In Proceedings of the 16th International Conference on Multimodal Interaction, pp. 80-81. 2014.
  (Show abstract) (Hide abstract) (PDF)
  

  We present MARCO, a hybrid, chess playing agent equipped with a custom-built robotic arm and a virtual agent’s face displaying emotions. MARCO was built to investigate the hypothesis that hybrid agents capable of displaying emotions make playing chess more personal and enjoyable. In addition, we aim to explore means of achieving emotional contagion between man and machine.

• Christian Becker-Asano, Eduardo Meneses, Nicolas Riesterer, Julien Hué, Christian Dornhege and Bernhard Nebel.
  The Hybrid Agent MARCO: A Multimodal Autonomous Robotic Chess Opponent.
  In Proceedings of the 2nd International Confernce on Human-Agent Interaction, pp. 173-176. 2014.
  (Show abstract) (Hide abstract) (PDF)
  

  We present MARCO, a hybrid, chess playing agent equipped with a custom-built robotic arm and a virtual agent’s face displaying emotions. MARCO was built to investigate the hypothesis that hybrid agents capable of displaying emo- tions make playing chess more personal and enjoyable. In addition, we aim to explore means of achieving emotional contagion between man and machine.

• Matthias Westphal and Julien Hué.
  A Concise Horn Theory for RCC8.
  In Proceedings of European Conference on Artificial Intelligence (ECAI'14). 2014.
  (Show abstract) (Hide abstract) (Translation Script; TAR.GZ)
  

  RCC8 is a well-known constraint language for expressing and reasoning about spatial knowledge. We state a simple and concise Horn theory for RCC8 analogous to the ORD-Horn theory for temporal reasoning. This theory allows for expressing RCC8 and retains tractability of the well-known Horn reduct of RCC8. Further, it is much more adequate for practical purposes in the area of logic programming and surpasses previous attempts.

• Julien Hué, Matthias Westphal and Stefan Wölfl.
  Towards a new semantic for Possibilistic Answer Sets.
  In Proceedings of Advances in Artificial Intelligence (KI'14). 2014.
  (Show abstract) (Hide abstract) (Springer Online; DOI) (DBLP)
  

  Possibilistic Answer Set Programming is an extension of the standard ASP framework that allows for attaching degrees of certainty to the rules in ASP programs. In the literature, several semantics for such PASP-programs have been presented, each of them having particular strengths and weaknesses. In this work we present a new semantics that employs so-called iota-answer sets, a solution concept introduced by Gebser et al.~(2009), in order to find solutions for standard ASP programs with odd cycles or auto-blocking rules. This is achieved by considering maximal subsets of a given ASP program for which answer sets exist. The main idea of our work is to integrate iota-semantics into the possibilistic framework in such a way that degrees of certainty are not only assigned to atoms mentioned in the answer sets, but also to the answer sets themselves. Our approach gives more satisfactory solutions and avoids counter-intuitive examples arising in the other approaches. We compare our approach to existing ones and present a translation into the standard ASP framework allowing the computation of solutions by existing tools.

• Matthias Westphal, Julien Hué and Stefan Wölfl.
  On the scope of Qualitative Constraint Calculi.
  In Proceedings of Advances in Artificial Intelligence (KI'14). 2014.
  (Show abstract) (Hide abstract) (Springer Online; DOI) (DBLP)
  

  Qualitative constraint calculi are a special kind of relation algebras defined by Ligozat and Renz for reasoning about binary constraints. Although this approach is known to be limited it has prevailed in the area of qualitative spatial and temporal reasoning. In this paper we revisit the definition of these calculi, contrast it with alternative approaches, and analyze general properties. Our results indicate that the concept of qualitative constraint calculi is both too narrow and too general: it disallows different approaches, but its setup already enables arbitrarily hard problems.

• Matthias Westphal, Julien Hué, Stefan Wölfl and Bernhard Nebel.
  Transition Constraints: A Study on the Computational Complexity of Qualitative Change.
  In Proceedings of International Joint Conference on Artificial Intelligence (IJCAI'13), pp. 1169-1175. 2013.
  (Show abstract) (Hide abstract) (Online; PDF) (DBLP)
  

  Many formalisms discussed in the literature on qualitative spatial reasoning are designed for expressing static spatial constraints only. However, dynamic situations arise in virtually all applications of these formalisms, which makes it necessary to study variants and extensions involving change. This paper presents a study on the computational complexity of qualitative change. More precisely, we discuss the reasoning task of finding a solution to a temporal sequence of static reasoning problems where this sequence is subject to additional transition constraints. Our focus is primarily on smoothness and continuity constraints: we show how such transitions can be defined as relations and expressed within qualitative constraint formalisms. Our results demonstrate that for point-based constraint formalisms the interesting fragments become NP-complete in the presence of continuity constraints, even if the satisfiability problem of its static descriptions is tractable.

• Matthias Westphal, Julien Hué and Stefan Wölfl.
  On the Propagation Strength of SAT Encodings for Qualitative Temporal Reasoning.
  In Proceedings of International Conference on Tools for Artificial Intelligence (ICTAI'13). 2013.
  (Show abstract) (Hide abstract) (PDF) (DOI) (DBLP) (Translation Script; TAR.GZ)
  

  Several studies in Qualitative Spatial and Temporal Reasoning discuss translations of the satisfiability problem on qualitative constraint languages into propositional SAT. Most of these encodings focus on compactness, while propagation strength is seldom discussed. In this work, we focus on temporal reasoning with the Point Algebra and Allen's Interval Algebra. We understand all encodings as a combination of propagation and search. We first give a systematic analysis of existing propagation approaches for these constraint languages. They are studied and ordered with respect to their propagation strength and refutation completeness for classes of input instances. Secondly, we discuss how existing encodings can be derived from such propagation approaches. We conclude our work with an empirical evaluation which shows that the older ORD-encoding by Nebel and Bürckert performs better than more recently suggested encodings.

• Julien Hué and Matthias Westphal.
  Revising Qualitative Constraint Network: Definition and Implementation.
  In Internationial Conference on Tools for Artificial Intelligence (ICTAI), pp. 548-555. 2012.
  (Show abstract) (Hide abstract) (PDF)
  

  Qualitative Spatial and Temporal Reasoning is a central topic in Artificial Intelligence. In particular, it is aimed at application scenarios dealing with uncertain information and thus needs to be able to handle dynamic beliefs. This makes merging and revision of qualitative information important topics. While merging has been studied extensively, revision which describes what is happening when one learns new information about a static world has been overlooked. In this paper, we propose to fill the gap by providing two revision operations for qualitative calculi. In order to implement these operations, we give algo- rithms for revision and analyze the computational complexity of these problems. Finally, we present an implementation of these algorithms based on a qualitative constraint solver and provide an experimental evaluation.

• Julien Hué, Matthias Westphal and Stefan Wölfl.
  An automatic decomposition method for qualitative spatial and temporal reasoning.
  In International Conference on Tools for Artificial Intelligence (ICTAI), pp. 588-595. 2012.
  (Show abstract) (Hide abstract) (PDF) (DBLP)
  

  Qualitative spatial and temporal reasoning is a research field that studies relational, constraint-based formalisms for representing, and reasoning about, spatial and temporal information. The standard approach for checking consistency is based on an exhaustive representation of possible configurations between three entities, the so-called composition tables. These tables, however, encode semantic background knowledge in a redundant way, which becomes a size and efficiency issue, when the composition table needs to be grounded as done in SAT encodings of problem instances. In this paper, we present a new framework that allows for decomposing composition tables into logically simpler parts, while preserving logical equivalence, e.g., the decomposition in start- and end-points for Allen’s Interval Calculus. We show that finding such decompositions is an NP-complete problem and present a SAT-based method to generate decompositions. Finally, we discuss the impact of our decomposition method on SAT encodings of problem instances, and present a reasoning system built on decompositions that compares favorably with state-of-the-art solvers.

• Matthias Westphal and Julien Hué.
  Nogoods in Qualitative Constraint-based Reasoning.
  In KI 2012: Advances in Artificial Intelligence (KI 2012), pp. 180-192. Springer-Verlag 2012.
  (Authors' preprint. The final publication is available at www.springerlink.com.).
  (Show abstract) (Hide abstract) (PDF)
  

  The prevalent method of increasing reasoning efficiency in the domain of qualitative constraint-based spatial and temporal reasoning is to use domain splitting based on so-called tractable subclasses. In this paper we analyze the application of nogood learning with restarts in combination with domain splitting. Previous results on nogood recording in the constraint satisfaction field feature learnt nogoods as a global constraint that allows for enforcing generalized arc consistency. We present an extension of such a technique capable of handling domain splitting, evaluate its benefits for qualitative constraint-based reasoning, and compare it with alternative approaches.

• Salem Benferhat, Julien Hué, Sylvain Lagrue and Julien Rossit.
  Merging Interval-Based Possibilistic Belief Bases.
  In International Conference on Scalable Uncertainty Management (SUM), pp. 447-458. 2012.
  (Show abstract) (Hide abstract) (PDF)
  

• Julien Hué, Mariette Sérayet, Pierre Drap, Odile Papini and Eric Würbel.
  Underwater Archaeological 3D Surveys Validation within the Removed Sets Framework.
  In Benchmarks and Applications of Spatial Reasoning (BASR), pp. 39-46. 2011.
  (Show abstract) (Hide abstract) (PDF)
  

Thomas Keller

• Florian Geißer, David Speck and Thomas Keller.
  Trial-based Heuristic Tree Search for MDPs with Factored Action Spaces.
  In Proceedings of the 13th Annual Symposium on Combinatorial Search (SoCS 2020), pp. 38-47. 2020.
  (Show abstract) (Hide abstract) (PDF)
  

  MDPs with factored action spaces, i.e. where actions are described as assignments to a set of action variables, allow reasoning over action variables instead of action states, yet most algorithms only consider a grounded action representation. This includes algorithms that are instantiations of the trial-based heuristic tree search (THTS) framework, such as AO* or UCT. To be able to reason over factored action spaces, we propose a generalisation of THTS where nodes that branch over all applicable actions are replaced with subtrees that consist of nodes that represent the decision for a single action variable. We show that many THTS algorithms retain their theoretical properties under the generalised framework, and show how to approximate any state-action heuristic to a heuristic for partial action assignments. This allows to guide a UCT variant that is able to create exponentially fewer nodes than the same algorithm that considers ground actions. An empirical evaluation on the benchmark set of the probabilistic track of the latest International Planning Competition validates the benefits of the approach.

• Florian Geißer, David Speck and Thomas Keller.
  An Analysis of the Probabilistic Track of the IPC 2018.
  In Proceedings of the ICAPS-2019 Workshop on the International Planning Competition (WIPC 2019), pp. 27-35. 2019.
  (Show abstract) (Hide abstract) (PDF)
  

  The International Planning Competition 2018 consisted of several tracks on classical, temporal and probabilistic planning. In this paper, we focus on the discrete MDP track of the probabilistic portion of the competition.

  We discuss the changes to the input language RDDL, which give rise to new challenges for planning systems, and analyze each of the eight competition domains separately and highlight unique properties. We demonstrate flaws of the used evaluation criterion, the IPC score, and discuss the need for optimal upper bounds. An evaluation of the three top-performers, including their post-competition versions, and a brief analysis of their performance highlights the strengths and weaknesses of the individual approaches.

• Thomas Keller, Florian Pommerening, Jendrik Seipp, Florian Geißer and Robert Mattmüller.
  State-dependent Cost Partitionings for Cartesian Abstractions in Classical Planning (Extended Abstract).
  In Proceedings of the 39th German Conference on Artificial Intelligence (KI 2016). 2016.
  (Show abstract) (Hide abstract) (PDF)
  

  Abstraction heuristics are a popular method to guide optimal search algorithms in classical planning. Cost partitionings allow to sum heuristic estimates admissibly by partitioning action costs among the abstractions. We introduce state-dependent cost partitionings which take context information of actions into account, and show that an optimal state-dependent cost partitioning dominates its state-independent counterpart. We demonstrate the potential of state-dependent cost partitionings with a state-dependent variant of the recently proposed saturated cost partitioning, and show that it can sometimes improve not only over its state-independent counterpart, but even over the optimal state-independent cost partitioning.

• Thomas Keller, Florian Pommerening, Jendrik Seipp, Florian Geißer and Robert Mattmüller.
  State-dependent Cost Partitionings for Cartesian Abstractions in Classical Planning.
  In Proceedings of the 25th International Joint Conference on Artificial Intelligence (IJCAI 2016). 2016.
  (Show abstract) (Hide abstract) (PDF)
  

  Abstraction heuristics are a popular method to guide optimal search algorithms in classical planning. Cost partitionings allow to sum heuristic estimates admissibly by distributing action costs among the heuristics. We introduce state-dependent cost partitionings which take context information of actions into account, and show that an optimal state-dependent cost partitioning dominates its state-independent counterpart. We demonstrate the potential of our idea with a state-dependent variant of the recently proposed saturated cost partitioning, and show that it has the potential to improve not only over its state-independent counterpart, but even over the optimal state-independent cost partitioning. Our empirical results give evidence that ignoring the context of actions in the computation of a cost partitioning leads to a significant loss of information.

• Florian Geißer, Thomas Keller and Robert Mattmüller.
  Abstractions for Planning with State-Dependent Action Costs.
  In Proceedings of the 26th International Conference on Automated Planning and Scheduling (ICAPS 2016). 2016.
  (Show abstract) (Hide abstract) (PDF)
  

  Extending the classical planning formalism with state-dependent action costs (SDAC) allows an up to exponentially more compact task encoding. Recent work proposed to use edge-valued multi-valued decision diagrams (EVMDDs) to represent cost functions, which allows to automatically detect and exhibit structure in cost functions and to make heuristic estimators accurately reflect SDAC. However, so far only the inadmissible additive heuristic has been considered in this context. In this paper, we define informative admissible abstraction heuristics which enable optimal planning with SDAC. We discuss how abstract cost values can be extracted from EVMDDs that represent concrete cost functions without adjusting them to the selected abstraction. Our theoretical analysis shows that this is efficiently possible for abstractions that are Cartesian or coarser. We adapt the counterexample-guided abstraction refinement approach to derive such abstractions. An empirical evaluation of the resulting heuristic shows that highly accurate values can be computed quickly.

• Florian Geißer, Thomas Keller and Robert Mattmüller.
  Delete Relaxations for Planning with State-Dependent Action Costs.
  In Proceedings of the 24th International Joint Conference on Artificial Intelligence (IJCAI 2015). 2015.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  Most work in planning focuses on tasks with state-independent or even uniform action costs. However, supporting state-dependent action costs admits a more compact representation of many tasks. We investigate how to solve such tasks using heuristic search, with a focus on delete-relaxation heuristics. We first define a generalization of the additive heuristic to such tasks and then discuss different ways of computing it via compilations to tasks with state-independent action costs and more directly by modifying the relaxed planning graph. We evaluate these approaches theoretically and present an implementation of the generalized additive heuristic for planning with state-dependent action costs. To our knowledge, this gives rise to the first approach able to handle even the hardest instances of the combinatorial Academic Advising domain from the International Probabilistic Planning Competition (IPPC) 2014.

• Florian Geißer, Thomas Keller and Robert Mattmüller.
  Delete Relaxations for Planning with State-Dependent Action Costs.
  In Proceedings of the 8th Annual Symposium on Combinatorial Search (SoCS 2015). 2015.
  Extended abstract of the IJCAI 2015 paper by the same name.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  Supporting state-dependent action costs in planning admits a more compact representation of many tasks. We generalize the additive heuristic and compute it by embedding decision-diagram representations of action cost functions into the RPG. We give a theoretical evaluation and present an implementation of the generalized additive heuristic. This allows us to handle even the hardest instances of the combinatorial Academic Advising domain from the IPPC 2014.

• Thomas Keller and Florian Geißer.
  Better Be Lucky Than Good: Exceeding Expectations in MDP Evaluation.
  In Proceedings of the 29th AAAI Conference on Artificial Intelligence (AAAI 2015). AAAI Press 2015.
  Erratum: On page 7, we mention that the results at IPPC would have differed by "-0.09", "+0.04" and "+0.05", which should read "-0.009", "+0.004" and "+0.005" instead.
  (Show abstract) (Hide abstract) (PDF)
  

  We introduce the MDP-Evaluation Stopping Problem, the optimization problem faced by participants of the International Probabilistic Planning Competition 2014 that focus on their own performance. It can be constructed as a meta-MDP where actions correspond to the application of a policy on a base-MDP, which is intractable in practice. Our theoretical analysis reveals that there are tractable special cases where the problem can be reduced to an optimal stopping problem. We derive approximate strategies of high quality by relaxing the general problem to an optimal stopping problem, and show both theoretically and experimentally that it not only pays off to pursue luck in the execution of the optimal policy, but that there are even cases where it is better to be lucky than good as the execution of a suboptimal base policy is part of an optimal strategy in the meta-MDP.

• Tim Schulte and Thomas Keller.
  Balancing Exploration and Exploitation in Classical Planning.
  In Proceedings of the Seventh Annual Symposium on Combinatorial Search (SoCS 2014) (SoCS 2014). 2014.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  Successful heuristic search planners for satisficing planning like FF or LAMA are usually based on one or more best first search techniques. Recent research has led to planners like Arvand, Roamer or Probe, where novel techniques like Monte-Carlo Random Walks extend the traditional exploitation-focused best first search by an exploration component. The UCT algorithm balances these contradictory incentives and has shown tremendous success in related areas of sequential decision making but has never been applied to classical planning yet. We make up for this shortcoming by applying the Trial-based Heuristic Tree Search framework to classical planning. We show how to model the best first search techniques Weighted A* and Greedy Best First Search with only three ingredients: action selection, initialization and backup function. Then we use THTS to derive four versions of the UCT algorithm that differ in the used backup functions. The experimental evaluation shows that our main algorithm, GreedyUCT*, outperforms all other algorithms presented in this paper, both in terms of coverage and quality.

• Andreas Hertle, Christian Dornhege, Thomas Keller, Robert Mattmüller, Manuela Ortlieb and Bernhard Nebel.
  An Experimental Comparison of Classical, FOND and Probabilistic Planning.
  In Proceedings of the 37th German Conference on Artificial Intelligence (KI 2014), pp. 297-308. Springer 2014.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  Domain-independent planning in general is broadly applicable to a wide range of tasks. Many formalisms exist that allow to describe different aspects of realistic problems. Which one is best is not clear as usually more expressiveness comes with a cost in planning time. Under the assumption that hard guarantees are not required, users are faced with a decision between multiple approaches. In this work we study the effect of nondeterministic action outcomes. As a generic model we use a probabilistic description in the form of Markov Decision Processes (MDPs). We define abstracting translations into a classical planning formalism and fully observable nondeterministic (FOND) planning. Our goal is to give insight into how state-of-the-art systems perform on different MDP planning domains. We evaluate those MDPs by running state-of-the-art planning systems on the abstracted formalisms.

• Florian Geißer, Thomas Keller and Robert Mattmüller.
  Past, Present, and Future: An Optimal Online Algorithm for Single-Player GDL-II Games.
  In Proceedings of the 21st European Conference on Artificial Intelligence (ECAI 2014), pp. 357-362. 2014.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  In General Game Playing, a player receives the rules of an unknown game and attempts to maximize his expected reward. Since 2011, the GDL-II rule language extension allows the formulation of nondeterministic and partially observable games. In this paper, we present an algorithm for such games, with a focus on the single-player case. Conceptually, at each stage, the proposed Norns algorithm distinguishes between the past, present and future steps of the game. More specifically, a belief state tree is used to simulate a potential past that leads to a present that is consistent with received observations. Unlike other related methods, our method is asymptotically optimal. Moreover, augmenting the belief state tree with iteratively improved probabilities speeds up the process over time significantly. As this allows a true picture of the present, we additionally present an optimal version of the well-known UCT algorithm for partially observable single-player games. Instead of performing hindsight optimization on a simplified, fully observable tree, the true future is simulated on an action-observation tree that takes partial observability into account. The expected reward estimates of applicable actions converge towards the true expected rewards even for moves that are only used to gather information. We prove that our algorithm is asymptotically optimal for single-player games and POMDPs and support our claim with an empirical evaluation.

• Thomas Keller and Malte Helmert.
  Trial-based Heuristic Tree Search for Finite Horizon MDPs.
  In Proceedings of the 1st Multidisciplinary Conference on Reinforcement Learning and Decision Making (RLDM 2013), pp. 101-105. 2013.
  Extended abstract of the ICAPS 2013 paper by the same name.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  Dynamic programming is a well-known approach for solving MDPs. In large state spaces, asynchronous versions like Real-Time Dynamic Programming (RTDP) have been applied successfully. If unfolded into equivalent trees, Monte-Carlo Tree Search algorithms are a valid alternative. UCT, the most popular representative, obtains good anytime behavior by guiding the search towards promising areas of the search tree and supporting non-admissible heuristics. The global Heuristic Search algorithm AO* finds optimal solutions for MDPs that can be represented as acyclic AND/OR graphs. Despite the differences, these approaches actually have much in common. We present the Trial-based Heuristic Tree Search (THTS) framework that subsumes these approaches and distinguishes them based on only five ingredients: heuristic function, backup function, action selection, outcome selection, and trial length. We describe the ingredients that model RTDP, AO* and UCT within this framework, and use THTS to combine attributes of these algorithms step by step in order to derive novel algorithms with superior theoretical properties. We merge Full Bellman and Monte-Carlo backup functions to Partial Bellman backups, and gain a function that both allows partial updates and a procedure that labels states when they are solved. DP-UCT combines attributes and theoretical properties from RTDP and UCT even though it differs from the latter only in the used Partial Bellman backups. Our main algorithm, UCT* adds a limited trial length to DP-UCT to inherit the global search behavior of AO*, which ensures that parts of the state space that are closer to the root are investigated more thoroughly. The experimental evaluation shows that both DP-UCT and UCT* are not only superior to UCT, but also outperform P ROST, the winner of the International Probabilistic Planning Competition (IPPC) 2011 on the benchmarks of IPPC 2011.

• Thomas Keller and Malte Helmert.
  Trial-based Heuristic Tree Search for Finite Horizon MDPs.
  In Proceedings of the 23rd International Conference on Automated Planning and Scheduling (ICAPS 2013), pp. 135-143. 2013.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  Dynamic programming is a well-known approach for solving MDPs. In large state spaces, asynchronous versions like Real-Time Dynamic Programming have been applied successfully. If unfolded into equivalent trees, Monte-Carlo Tree Search algorithms are a valid alternative. UCT, the most popular representative, obtains good anytime behavior by guiding the search towards promising areas of the search tree. The Heuristic Search algorithm AO∗ finds optimal solutions for MDPs that can be represented as acyclic AND/OR graphs.

  We introduce a common framework, Trial-based Heuristic Tree Search, that subsumes these approaches and distinguishes them based on five ingredients: heuristic function, backup function, action selection, outcome selection, and trial length. Using this framework, we describe three new algorithms which mix these ingredients in novel ways in an attempt to combine their different strengths. Our evaluation shows that two of our algorithms not only provide superior theoretical properties to UCT, but also outperform state-of-the-art approaches experimentally.

• Thomas Keller and Patrick Eyerich.
  PROST: Probabilistic Planning Based on UCT.
  In Proceedings of the 22nd International Conference on Automated Planning and Scheduling (ICAPS 2012), pp. 119-127. 2012.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  We present PROST, a probabilistic planning system that is based on the UCT algorithm by Kocsis and Szepesvari (2006), which has been applied successfully to many areas of planning and acting under uncertainty. The objective of this paper is to show the application of UCT to domain-independent probabilistic planning, an area it had not been applied to before. We furthermore present several enhancements to the algorithm, including a method that is able to drastically reduce the branching factor by identifying superfluous actions. We show how search depth limitation leads to a more thoroughly investigated search space in parts that are influential on the quality of a policy, and present a sound and polynomially computable detection of reward locks, states that correspond to, e.g., dead ends or goals. We describe a general Q-value initialization for unvisited nodes in the search tree that circumvents the initial random walks inherent to UCT, and leads to a faster convergence on average. We demonstrate the significant influence of the enhancements by providing a comparison on the IPPC benchmark domains.

• Johannes Löhr, Patrick Eyerich, Thomas Keller and Bernhard Nebel.
  A Planning Based Framework for Controlling Hybrid Systems.
  In Proceedings of the 22nd International Conference on Automated Planning and Scheduling (ICAPS 2012). 2012.
  (Show abstract) (Hide abstract) (PDF)
  

  The control of dynamic systems, which aims to minimize the deviation of state variables from reference values in a contin- uous state space, is a central domain of cybernetics and con- trol theory. The objective of action planning is to find feasible state trajectories in a discrete state space from an initial state to a state satisfying the goal conditions, which in principle ad- dresses the same issue on a more abstract level. We combine these approaches to switch between dynamic system charac- teristics on the fly, and to generate control input sequences that affect both discrete and continuous state variables. Our approach (called Domain Predictive Control) is applicable to hybrid systems with linear dynamics and discretizable inputs.

• Andreas Hertle, Christian Dornhege, Thomas Keller and Bernhard Nebel.
  Planning with Semantic Attachments: An Object-Oriented View.
  In Proceedings of the European Conference on Artificial Intelligence (ECAI). 2012.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  In recent years, domain-independent planning has been applied to a rising number of real-world applications. Usually, the description language of choice is PDDL. However, PDDL is not suited to model all challenges imposed by real-world applications. Dornhege et al. proposed semantic attachments to allow the computation of Boolean fluents by external processes called modules during planning. To acquire state information from the planning system a module developer must perform manual requests through a callback interface which is both inefficient and error-prone. In this paper, we present the Object-oriented Planning Language OPL, which incorporates the structure and advantages of modern object-oriented programming languages. We demonstrate how a domain-specific module interface that allows to directly access the planner state using object member functions is automatically gen- erated from an OPL planning task. The generated domain-specific interface allows for a safe and less error-prone implementation of modules. We show experimentally that this interface is more efficient than the PDDL-based module interface of TFD/M.

• Danijel Skocaj, Matej Kristan, Alen Vrecko, Marko Mahnic, Miroslav Janicek, Geert-Jan M. Kruijff, Marc Hanheide, Nick Hawes, Thomas Keller, Michael Zillich and Kai Zhou.
  A system for interactive learning in dialogue with a tutor.
  In Proceedings of the IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS 2011). 2011.
  (Show abstract) (Hide abstract) (PDF)
  

  In this paper we present representations and mechanisms that facilitate continuous learning of visual concepts in dialogue with a tutor and show the implemented robot system. We present how beliefs about the world are created by processing visual and linguistic information and show how they are used for planning system behaviour with the aim at satisfying its internal drive -- to extend its knowledge. The system facilitates different kinds of learning initiated by the human tutor or by the system itself. We demonstrate these principles in the case of learning about object colours and basic shapes.

• Thomas Keller and Patrick Eyerich.
  A Polynomial All Outcome Determinization for Probabilistic Planning.
  In Fahiem Bacchus, Carmel Domshlak, Stefan Edelkamp and Malte Helmert (eds.), Proceedings of the 21th International Conference on Automated Planning and Scheduling (ICAPS 2011), pp. 331-334. AAAI Press 2011.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  Most predominant approaches in probabilistic planning utilize techniques from the more thoroughly investigated field of classical planning by determinizing the problem at hand. In this paper, we present a method to map probabilistic operators to an equivalent set of probabilistic operators in a novel normal form, requiring polynomial time and space. From this, we directly derive a determinization which can be used for, \eg, replanning strategies incorporating a classical planning system. Unlike previously described all outcome determinizations, the number of deterministic operators is not exponentially but polynomially bounded in the number of parallel probabilistic effects, enabling the use of more sophisticated determinization-based techniques in the future.

• Thomas Keller, Patrick Eyerich and Bernhard Nebel.
  Task Planning for an Autonomous Service Robot.
  In Rüdiger Dillmann, Jürgen Beyerer, Uwe Hanebeck and Tanja Schultz (eds.), Proceedings on the 33rd Annual German Conference on Artificial Intelligence (KI 2010), pp. 358-365. Springer-Verlag 2010.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  In the DESIRE project an autonomous robot capable of performing service tasks in a typical kitchen environment has been developed. The overall system consists of various loosely coupled subcomponents providing particular features like manipulating objects or recognizing and interacting with humans. To bring all these subcomponents together to act as monolithic system, a high-performance planning system has been implemented. In this paper, we present this system’s basic architecture and some advanced extensions necessary to cope with the various challenges arising in dynamic and uncertain environments like those a real world service robot is usually faced with.

• Patrick Eyerich, Thomas Keller and Malte Helmert.
  High-Quality Policies for the Canadian Traveler's Problem.
  In Maria Fox and David Poole (eds.), Proceedings of the Twenty-Fourth AAAI Conference on Artificial Intelligence (AAAI 2010), pp. 51-58. AAAI Press 2010.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  We consider the stochastic variant of the Canadian Traveler's Problem, a path planning problem where adverse weather can cause some roads to be untraversable. The agent does not initially know which roads can be used. However, it knows a probability distribution for the weather, and it can observe the status of roads incident to its location. The objective is to find a policy with low expected travel cost.

  We introduce and compare several algorithms for the stochastic CTP. Unlike the optimistic approach most commonly considered in the literature, the new approaches we propose take uncertainty into account explicitly. We show that this property enables them to generate policies of much higher quality than the optimistic one, both theoretically and experimentally.

• Patrick Eyerich, Thomas Keller and Malte Helmert.
  High-Quality Policies for the Canadian Traveler's Problem (Extended Abstract).
  In Ariel Felner and Nathan Sturtevant (eds.), Proceedings of the Third Annual Symposium on Combinatorial Search (SoCS 2010), pp. 147-148. AAAI Press 2010.
  Extended abstract of the AAAI paper by the same name.
  (PDF)
  

• Moritz Göbelbecker, Thomas Keller, Patrick Eyerich, Michael Brenner and Bernhard Nebel.
  Coming Up with Good Excuses: What To Do When No Plan Can be Found.
  In Ronen Brafman, Héctor Geffner, Jörg Hoffmann and Henry Kautz (eds.), Proceedings of the 20th International Conference on Automated Planning and Scheduling (ICAPS 2010), pp. 81-88. AAAI Press 2010.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  When using a planner-based agent architecture, many things can go wrong. First and foremost, an agent might fail to execute one of the planned actions for some reasons. Even more annoying, however, is a situation where the agent is incompetent, i.e., unable to come up with a plan. This might be due to the fact that there are principal reasons that prohibit a successful plan or simply because the task's description is incomplete or incorrect. In either case, an explanation for such a failure would be very helpful. We will address this problem and provide a formalization of coming up with excuses for not being able to find a plan. Based on that, we will present an algorithm that is able to find excuses and demonstrate that such excuses can be found in practical settings in reasonable time.

• Patrick Eyerich, Thomas Keller and Malte Helmert.
  High-Quality Policies for the Canadian Traveler's Problem.
  In Proceedings of the ICAPS-2010 Workshop on Planning and Scheduling Under Uncertainty. 2010.
  Superseded by the AAAI 2010 paper by the same name.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  We consider the stochastic variant of the Canadian Traveler's Problem, a path planning problem where adverse weather can cause some roads to be untraversable. The agent does not initially know which roads can be used. However, it knows a probability distribution for the weather, and it can observe the status of roads incident to its location. The objective is to find a policy with low expected travel cost.

  We introduce and compare several algorithms for the stochastic CTP. Unlike the optimistic approach most commonly considered in the literature, the new approaches we propose take uncertainty into account explicitly. We show that this property enables them to generate policies of much higher quality than the optimistic one, both theoretically and experimentally.

• Patrick Eyerich, Thomas Keller and Bernhard Nebel.
  Combining Action and Motion Planning via Semantic Attachments.
  In Proceedings of the Workshop on Combining Action and Motion Planning at ICAPS 2010 (CAMP 2010), p. 19. 2010.
  Extended Abstract.
  (PDF) (BIB)
  

• Christian Dornhege, Patrick Eyerich, Thomas Keller, Sebastian Trüg, Michael Brenner and Bernhard Nebel.
  Semantic Attachments for Domain-Independent Planning Systems.
  In Proceedings of the 19th International Conference on Automated Planning and Scheduling (ICAPS 2009), pp. 114-121. AAAI Press 2009.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  Solving real-world problems using symbolic planning often requires a simplified formulation of the original problem, since certain subproblems cannot be represented at all or only in a way leading to inefficiency. For example, manipulation planning may appear as a subproblem in a robotic planning context or a packing problem can be part of a logistics task. In this paper we propose an extension of PDDL for specifying semantic attachments. This allows the evaluation of grounded predicates as well as the change of fluents by externally specified functions. Furthermore, we describe a general schema of integrating semantic attachments into a forward-chaining planner and report on our experience of adding this extension to the planners FF and Temporal Fast Downward. Finally, we present some preliminary experiments using semantic attachments.

• Thomas Keller and Sebastian Kupferschmid.
  Automatic Bidding for the Game of Skat.
  In Andreas R. Dengel, Karsten Berns, Thomas M. Breuel, Frank Bomarius and Thomas R. Roth-Berghofer (eds.), Proceedings of the 31st Annual German Conference on Artificial Intelligence (KI 2008), pp. 95-102. Springer-Verlag 2008.
  (Show abstract) (Hide abstract) (BIB) (PDF)
  

  In recent years, researchers started to study the game of Skat. The strength of existing Skat playing programs is definitely the card play phase. The bidding phase, however, was treated quite poorly so far. This is a severe drawback since bidding abilities influence the overall playing performance drastically. In this paper we present a powerful bidding engine which is based on a k-nearest neighbor algorithm.

Alexander Kleiner

• Christian Dornhege, Alexander Kleiner, Andreas Hertle and Andreas Kolling.
  Multirobot Coverage Search in Three Dimensions.
  Journal of Field Robotics 33 (4), pp. 537-558. 2016.
  (Show abstract) (Hide abstract) (PDF)
  

  Searching for objects and observing parts of a known environment efficiently is a fundamental problem in many real-world robotic applications, e.g., household robots searching for objects, inspection robots searching for leaking pipelines, and rescue robots searching for survivors after a disaster. We consider the problem of identifying and planning sequences of sensor locations from which robot sensors can observe and cover complex three-dimensional (3D) environments while traveling only short distances. Our approach is based on sampling and ranking a large number of sensor locations for a 3D environment represented by an OctoMap. The visible area from these sensor locations induces a minimal partition of the 3D environment that we exploit for planning sequences of sensor locations with short travel times efficiently. We present multiple planning algorithms designed for single robots and for multirobot teams. These algorithms include variants that are greedy, optimal, or based on decomposing the planning problem into a set cover and traveling salesman problem. We evaluated and compared these algorithms empirically in simulation and real-world robot experiments with up to four robots. Our results demonstrate that, despite the intractability of the overall problem, computing and executing effective solutions for multirobot coverage search in real 3D environments is feasible and ready for real-world applications.

• Dali Sun, Alexander Kleiner and Bernhard Nebel.
  Behavior-based Multi-Robot Collision Avoidance.
  In Proceedings of the IEEE International Conference on Robotics and Automation (ICRA-14), pp. 1668-1673. 2014.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  Autonomous robot teams that simultaneously disatch transportation tasks are playing a more and more important role in the industry. In this paper we consider the multi-robot motion planning problem in large robot teams and present a decoupled approach by combining decentralized path planning methods and swarm technologies. Instead of a central coordination, a proper behavior which is directly selected according to the context is used by the robot to keep cooperating with others and to resolve path collisions. We show experimentally that the quality of solutions and the scalability of our method are significantly better than those of conventional decoupled path planning methods. Furthermore, compared to conventional swarm approaches, our method can be widely applied in large-scale environments.

• Christian Dornhege, Alexander Kleiner and Andreas Kolling.
  Coverage Search in 3D.
  In Proceedings of the Symposium on Safety Security and Rescue Robotics (SSRR). 2013.
  (PDF) (BIB)
  

• Christian Dornhege and Alexander Kleiner.
  A Frontier-Void-Based Approach for Autonomous Exploration in 3D.
  Advanced Robotics 27 (6). 2013.
  (BIB)
  

• Christian Dornhege and Alexander Kleiner.
  A Frontier-Void-Based Approach for Autonomous Exploration in 3D.
  In Proceedings of the IEEE International Symposium on Safety, Security and Rescue Robotics (SSRR). 2011.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  We consider the problem of an autonomous robot searching for objects in unknown 3d space. Similar to the well known frontier-based exploration in 2d, the problem is to determine a minimal sequence of sensor viewpoints until the entire search space has been explored. We introduce a novel approach that combines the two concepts of voids, which are unexplored volumes in 3d, and frontiers, which are regions on the boundary between voids and explored space. Our approach has been evaluated on a mobile platform equipped with a manipulator searching for victims in a simulated USAR setup. First results indicate the real-world capability and search efficiency of the proposed method.

• Alexander Kleiner, Dali Sun and D. Meyer-Delius.
  ARMO: Adaptive Road Map Optimization for Large Robot Teams.
  In Proc. of the IEEE/RSJ Int. Conf. on Intelligent Robots and Systems (IROS). 2011.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  Autonomous robot teams that simultaneously dispatch transportation tasks are playing more and more an important role in present logistic centers and manufacturing plants. In this paper we consider the problem of robot motion planning for large robot teams in the industrial domain. We present adaptive road map optimization (ARMO) that is capable of adapting the road map in real time whenever the environment has changed. Based on linear programming, ARMO computes an optimal road map according to current environmental constraints (including human whereabouts) and the current demand for transportation tasks from loading stations in the plant. For detecting dynamic changes, the environment is describe by a grid map augmented with a hidden Markov model (HMM). We show experimentally that ARMO outperforms decoupled planning in terms of computation time and time needed for task completion.

• Alexander Kleiner, A. Kolling, K. Sycara and M. Lewis.
  Hierarchical Visibility for Guaranteed Search in Large-Scale Outdoor Terrain.
  Journal of Autonomous Agents and Multi-Agent Systems (JAAMAS). 2011.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  Searching for moving targets in large environments is a challenging task that is relevant in several problem domains, such as capturing an invader in a camp, guarding security facilities, and searching for victims in large-scale search and rescue scenarios. The guaranteed search problem is to coordinate the search of a team of agents to guarantee the discovery of all targets. In this paper we present a self-contained solution to this problem in 2.5D real-world domains represented by digital elevation models (DEMs). We introduce hierarchical sampling on DEMs for selecting heuristically the close to minimal set of locations from which the entire surface of the DEM can be guarded. Locations are utilized to form a search graph on which search strategies for mobile agents are computed. For these strategies schedules are derived which include agent paths that are directly executable in the terrain. Presented experimental results demonstrate the performance of the method. The practical feasibility of our approach has been validated during a field experiment at the Gascola robot training site where teams of humans equipped with iPads successfully searched for adversarial and omniscient evaders. The field demonstration is the largest-scale implementation of a guaranteed search algorithm to date.

• Q. Hamp, L. Reindl and Alexander Kleiner.
  Lessons Learned from German Research for USAR.
  In Proc. of the IEEE Int. Workshop on Safety, Security and Rescue Robotics (SSRR). 2011.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  We present lessons learned in USAR research within the framework of the German research project I-LOV. After three years of development first field tests have been carried out by professionals such as the Rapid Deployment Unit for Salvage Operations Abroad (SEEBA). We present results from evaluating search teams in simulated USAR scenarios equipped with newly developed technical search means and digital data input terminals developed in the I- LOV project. In particular, the bioradar, a ground-penetrating radar system for the detection of humanoid movements, a semi-active video probe for rubble pile exploration of more than 10 m length, and the decision support system FRIEDAA were evaluated and compared with conventional search methods. Results of this evaluation indicate that the developed technologies foster advantages in USAR, which are discussed in this paper.

• Alexander Kleiner, Bernhard Nebel and V.A. Ziparo.
  A Mechanism for Dynamic Ride Sharing based on Parallel Auctions.
  In Proc. of the 22th International Joint Conference on Artificial Intelligence (IJCAI). 2011.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  Car pollution is one of the major causes of green- house emissions, and traffic congestion is rapidly becoming a social plague. Dynamic Ride Sharing (DRS) systems have the potential to mitigate this problem by computing plans for car drivers, e.g. commuters, allowing them to share their rides. Ex- isting efforts in DRS are suffering from the problem that participants are abandoning the system after repeatedly failing to get a shared ride. In this paper we present an incentive compatible DRS solution based on auctions. While existing DRS systems are mainly focusing on fixed assignments that minimize the totally travelled distance, the presented approach is adaptive to individual preferences of the participants. Furthermore, our system allows to tradeoff the minimization of Vehicle Kilometers Travelled (VKT) with the overall probability of successful ride-shares, which is an important feature when bootstrapping the system. To the best of our knowledge, we are the first to present a DRS solution based on auctions using a sealed-bid second price scheme.

• D. Meyer-Delius, M. Beinhofer, Alexander Kleiner and W. Burgard.
  Reducing the Ambiguity in the Environment by Placing Artificial Landmarks to Improve Mobile Robot Localization.
  In Proc. of the IEEE Int. Conf. on Robotics and Automation (ICRA). 2011.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  Robust and reliable localization is a fundamental prerequisite for successful robot navigation. Although there exist many solutions to the localization problem, environments can be inherently ambiguous so that different robot locations cannot to be distinguished using the sensors of the robot. This is particularly critical in comercial environments, for example in warehouses, containing long corridors and symmetric structures. This ambiguity makes localization approaches more likely to diverge or even prevent the pose of the robot from being uniquely estimated at all. In this paper we propose the utilization of artificial landmarks to reduce the inherent ambiguity in the environment, and present an efficient, localization-oriented approach to landmark placement. Our approach provides us with both the location and number of landmarks to be placed in order to improve the localization performance of the robot in a given environment. Experimental results show that by intelligently placing the landmarks we can improve the localization performance of the robot.

• A. Kolling, Alexander Kleiner, M. Lewis and K. Sycara.
  Computing and Executing Strategies for Moving Target Search.
  In Proc. of the IEEE Int. Conf. on Robotics and Automation (ICRA). 2011.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  We address the problem of searching for moving targets in large outdoor environments represented by height maps. To solve the problem we present a complete system that computes from an annotated height map a graph representation and search strategies based on worst-case assumptions about all targets. These strategies are then used to compute a schedule and task assignment for all agents. We improve the graph construction from previous work and for the first time present a method that computes a schedule to minimize the execution time. For this we consider travel times of agents determined by a path planner on the height map. We demonstrate the entire system in a real environment with an area of 700,000m2 in which eight human agents search for two intruders using mobile computing devices (iPads). To the best of our knowledge this is the first demonstration of a search system applied to such a large environment.

• R. Kümmerle, B. Steder, Christian Dornhege, Alexander Kleiner, G. Grisetti and W. Burgard.
  Large Scale Graph-based SLAM using Aerial Images as Prior Information.
  Autonomous Robots 30, pp. 25-39. 2011.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  The problem of learning a map with a mobile robot has been intensively studied in the past and is usually referred to as the simultaneous localization and mapping (SLAM) problem. However, most existing solutions to the SLAM problem learn the maps from scratch and have no means for incorporating prior information. In this paper, we present a novel SLAM approach that achieves global consistency by utilizing publicly accessible aerial photographs as prior information. It inserts correspondences found between stereo and three-dimensional range data and the aerial images as constraints into a graph-based formulation of the SLAM problem. We evaluate our algorithm based on large real-world datasets acquired even in mixed in- and outdoor environments by comparing the global accuracy with state-of-the-art SLAM approaches and GPS. The experimental results demonstrate that the maps acquired with our method show increased global consistency.

• Wei Mou and Alexander Kleiner.
  Online Learning Terrain Classification for Adaptive Velocity Control.
  In Proceedings of the IEEE Int. Workshop on Safety, Security and Rescue Robotics (SSRR 2010). 2010.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  Safe teleoperation during critical missions, such as urban search and rescue and bomb disposal, requires careful velocity control when different types of terrain are found in the scenario. This can particularly be challenging when mission time is limited and the operator’s field of view affected. This paper presents a method for online adapting robot velocities according to the terrain classification from vision and laser readings. The classifier adapts itself to illumination variations, and can be improved online given feedback from the operator.

• Andreas Kolling, Alexander Kleiner, Michael Lewis and Katia Sycara.
  Solving Pursuit-Evasion Problems on Height Maps.
  In IEEE International Conference on Robotics and Automation (ICRA 2010) Workshop: Search and Pursuit/Evasion in the Physical World: Efficiency, Scalability, and Guarantees (WSPE ICRA 2010). 2010.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  In this paper we present an approach for a pursuit-evasion problem that considers a 2.5d environment represented by a height map. Such a representation is particularly suitable for large-scale outdoor pursuit-evasion. By allowing height information we not only capture some aspects of 3d visibility but can also consider target heights. In our approach we construct a graph representation of the environment by sampling points and their detection sets which extend the usual notion of visibility. Once a graph is constructed we compute strategies on this graph using a modification of previous work on graph-searching. This strategy is converted into robot paths that are planned on the height map by classifying the terrain appropriately. In experiments we investigate the performance of our approach and provide examples including a map of a small village with surrounding hills and a sample map with multiple loops and elevation plateaus. Experiments are carried out with varying sensing ranges as well as target and sensor heights. To the best of our knowledge the presented approach is the first viable solution to 2.5d pursuit-evasion with height maps.

• Daniel Maier and Alexander Kleiner.
  Improved GPS Sensor Model for Mobile Robots in Urban Terrain.
  In IEEE International Conference on Robotics and Automation (ICRA 2010), pp. 4385-4390. 2010.
  (Video).
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  Autonomous robot navigation in outdoor scenarios gains increasing importance in various growing application areas. Whereas in non-urban domains such as deserts the problem of successful GPS-based navigation appears to be almost solved, navigation in urban domains particularly in the close vicinity of buildings is still a challenging problem. In such situations GPS accuracy significantly drops down due to multiple signal reflections with larger objects causing the so-called multipath error. In this paper we contribute a novel approach for incorporating multi- path errors into the conventional GPS sensor model by analyzing environmental structures from online generated point clouds. The approach has been validated by experimental results conducted with an all- terrain robot operating in scenarios requiring close- to-building navigation. Presented results show that positioning accuracy can significantly be improved within urban domains.

• Alexander Kleiner and Christian Dornhege.
  Mapping for the Support of First Responders in Critical Domains.
  Journal of Intelligent and Robotic Systems (JINT), pp. 1-29. 2010.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  In critical domains such as urban search and rescue (USAR), and bomb disposal, the deployment of teleoperated robots is essential to reduce the risk of first responder personnel. Teleoperation is a difficult task, particularly when controlling robots from an isolated safety zone. In general, the operator has to solve simultaneously the problems of mission planning, target identification, robot navigation, and robot control. We introduce a system to support teleoperated navigation with real-time mapping consisting of a two-step scan matching method that re-considers data associations during the search. The algorithm processes data from laser range finder and gyroscope only, thereby it is independent from the robot platform. Furthermore, we introduce a user-guided procedure for improving the global consistency of maps generated by the scan matcher. Globally consistent maps are computed by a graph-based maximum likelihood method that is biased by localizing crucial parts of the scan matcher trajectory on a prior given geo-tiff image. The approach has been implemented as an embedded system and extensively tested on robot platforms designed for teleoperation in critical situations, such as bomb disposal. Furthermore, the system was evaluated in a test maze by first responders during the Disaster City event in Texas 2008.

• Sören Schwertfeger, Adam Jacoff, Chris Scrapper, Johannes Pellenz and Alexander Kleiner.
  Evaluation of Maps using Fixed Shapes: The Fiducial Map Metric.
  In Proc. of the Int. Workshop on Performance Metrics for Intelligent Systems (PerMIS), pp. 344-351. NIST 2010.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  Mapping is an important task for mobile robots. Assessing the quality of those maps is an open topic. A new approach on map evaluation is presented here. It makes use of artificial objects placed in the environment named "Fiducials". Using the known ground-truth positions and the positions of the fiducials identified in the map, a number of quality attributes can be assigned to that map. Depending on the application domain those attributes are weighed to compute a final score. During the 2010 NIST Response Robot Evaluation Exercise at Disaster City an area was populated with fiducials and different mapping runs were performed. The maps generated there are assessed in this paper demonstrating the Fiducial approach. Finally this map scoring algorithm is compared to other approaches found in literature.

• A. Kolling, Alexander Kleiner, M. Lewis and and K. Sycara.
  Pursuit-Evasion in 2.5d based on Team-Visibility.
  In Proceedings of the IEEE/RSJ Int. Conf. on Intelligent Robots and Systems (IROS 2010), pp. 4610-4616. 2010.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  In this paper we present an approach for a pursuit-evasion problem that considers a 2.5d environment represented by a height map. Such a representation is particularly suitable for large-scale outdoor pursuit-evasion, captures some aspects of 3d visibility and can include target heights. In our approach we construct a graph representation of the environment by sampling points and computing detection sets, an extended notion of visibility. Moreover, the constructed graph captures overlaps of detection sets allowing for a coordinated team-based clearing of the environment with robots that move to the sampled points. Once a graph is constructed we compute strategies on it utilizing previous work on graph-searching. This is converted into robot paths that are planned on the height map by classifying the terrain appropriately. In experiments we investigate the performance of our approach and provide examples including a sample map with multiple loops and elevation plateaus and two realistic maps, one of a village and one of a mountain range. To the best of our knowledge the presented approach is the first viable solution to 2.5d pursuit-evasion with height maps.

• Dali Sun, Alexander Kleiner and and C. Schindelhauer.
  Decentralized Hash Tables For Mobile Robot Teams Solving Intra-Logistics Tasks.
  In Proceedings of the 9th Int. Joint Conf. on Autonomous Agents and Multiagent Systems (AAMAS 2010), pp. 923-930. 2010.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  Although a remarkably high degree of automation has been reached in production and intra-logistics nowadays, human labor is still used for transportation using handcarts and forklifts. High labor cost and risk of injury are the undesirable consequences. Alternative approaches in automated warehouses are fixed installed conveyors installed either overhead or floor-based. The drawback of such solutions is the lack of flexibility, which is necessary when the production lines of the company change. Then, such an installation has to be re-built. In this paper, we propose a novel approach of decentralized teams of autonomous robots performing intra-logistics tasks using distributed algorithms. Centralized solutions suffer from limited scalability and have a single point of failure. The task is to transport material between stations keeping the communication network structure intact and most importantly, to facilitate a fair distribution of robots among loading stations. Our approach is motivated by strategies from peer-to-peer-networks and mobile ad-hoc networks. In particular we use an adapted version of distributed heterogeneous hash tables (DHHT) for distributing the tasks and localized communication. Experimental results presented in this paper show that our method reaches a fair distribution of robots over loading stations.

• Alexander Kleiner, Chris Scrapper and Adam Jacoff.
  RoboCupRescue Interleague Challenge 2009: Bridging the gap between Simulation and Reality.
  In Proceedings of the Int. Workshop on Performance Metrics for Intelligent Systems (Permis 2009), pp. 123-129. NIST 2009.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  Teleoperation is a difficult task, particularly when controlling robots from an isolated operator station. In general, the operator has to solve nearly blindly the problems of mission planning, target identification, robot navigation, and robot control at the same time. The goal of the proposed system is to support teleoperated navigation with real-time mapping. We present a novel scan matching technique that re-considers data associations during the search, enabling robust pose estimation even under varying roll and pitch angle of the robot enabling mapping on rough terrain. The approach has been implemented as an embedded system and extensively tested on robot platforms designed for teleoperation in critical situations, such as bomb disposal. Furthermore, the system has been evaluated in a test maze by first responders during the Disaster City event in Texas 2008. Finally, experiments conducted within different environments show that the system yields comparably accurate maps in real-time when compared to higher sophisticated offline methods, such as Rao-Blackwellized SLAM.

• Alexander Kleiner and Christian Dornhege.
  Operator-Assistive Mapping in Harsh Environments.
  In IEEE International Workshop on Safety, Security and Rescue Robotics (SSRR 2009), pp. 1-6. 2009.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  Teleoperation is a difficult task, particularly when controlling robots from an isolated operator station. In general, the operator has to solve nearly blindly the problems of mission planning, target identification, robot navigation, and robot control at the same time. The goal of the proposed system is to support teleoperated navigation with real-time mapping. We present a novel scan matching technique that re-considers data associations during the search, enabling robust pose estimation even under varying roll and pitch angle of the robot enabling mapping on rough terrain. The approach has been implemented as an embedded system and extensively tested on robot platforms designed for teleoperation in critical situations, such as bomb disposal. Furthermore, the system has been evaluated in a test maze by first responders during the Disaster City event in Texas 2008. Finally, experiments conducted within different environments show that the system yields comparably accurate maps in real-time when compared to higher sophisticated offline methods, such as Rao-Blackwellized SLAM.

• Rainer Kümmerle, Bastian Steder, Christian Dornhege, Michael Ruhnke, Giorgio Grisetti, Cyrill Stachniss and Alexander Kleiner.
  On Measuring the Accuracy of SLAM Algorithms.
  Autonomous Robots 27 (4), pp. 387-407. 2009.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  In this paper, we address the problem of creating an objective benchmark for evaluating SLAM approaches. We propose a framework for analyzing the results of a SLAM approach based on a metric for measuring the error of the corrected trajectory. This metric uses only relative relations between poses and does not rely on a global reference frame. This overcomes serious shortcomings of approaches using a global reference frame to compute the error. Our method furthermore allows us to compare SLAM approaches that use different estimation techniques or different sensor modalities since all computations are made based on the corrected trajectory of the robot.

  We provide sets of relative relations needed to compute our metric for an extensive set of datasets frequently used in the robotics community. The relations have been obtained by manually matching laser-range observations to avoid the errors caused by matching algorithms. Our benchmark framework allows the user to easily analyze and objectively compare different SLAM approaches.

• Wolfram Burgard, Cyrill Stachniss, Giorgio Grisetti, Bastian Steder, Rainer Kümmerle, Christian Dornhege, Michael Ruhnke, Alexander Kleiner and Juan D. Tardos.
  A Comparison of SLAM Algorithms Based on a Graph of Relations.
  In Proceedings of the 2009 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS 2009), pp. 2089-2095. IEEE 2009.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  In this paper, we address the problem of creating an objective benchmark for comparing SLAM approaches. We propose a framework for analyzing the results of SLAM approaches based on a metric for measuring the error of the corrected trajectory. The metric uses only relative relations between poses and does not rely on a global reference frame. The idea is related to graph-based SLAM approaches in the sense that it considers the energy needed to deform the trajectory estimated by a SLAM approach to the ground truth trajectory. Our method enables us to compare SLAM approaches that use different estimation techniques or different sensor modalities since all computations are made based on the corrected trajectory of the robot. We provide sets of relative relations needed to compute our metric for an extensive set of datasets frequently used in the SLAM community. The relations have been obtained by manually matching laser-range observations. We believe that our benchmarking framework allows the user an easy analysis and objective comparisons between different SLAM approaches.

• Rainer Kümmerle, Bastian Steder, Christian Dornhege, Alexander Kleiner, Giorgio Grisetti and Wolfram Burgard.
  Large Scale Graph-based SLAM using Aerial Images as Prior Information.
  In Proceedings of 2009 Robotics: Science and Systems (RSS 2009). 2009.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  To effectively navigate in their environments and accurately reach their target locations, mobile robots require a globally consistent map of the environment. The problem of learning a map with a mobile robot has been intensively studied in the past and is usually referred to as the simultaneous localization and mapping (SLAM) problem. However, existing solutions to the SLAM problem typically rely on loop-closures to obtain global consistency and do not exploit prior information even if it is available. In this paper, we present a novel SLAM approach that achieves global consistency by utilizing publicly accessible aerial photographs as prior information. Our approach inserts correspondences found between three-dimensional laser range scans and the aerial image as constraints into a graph-based formulation of the SLAM problem. We evaluate our algorithm based on large real-world datasets acquired in a mixed in- and outdoor environment by comparing the global accuracy with state-of-the-art SLAM approaches and GPS. The experimental results demonstrate that the maps acquired with our method show increased global consistency.

• Dali Sun, Alexander Kleiner and T. M. Wendt.
  Multi-Robot Range-Only SLAM by Active Sensor Nodes for Urban Search and Rescue.
  In Robocup 2008: Robot Soccer World Cup XII, pp. 318-330. Springer 2009.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  To jointly map an unknown environment with a team of autonomous robots is a challenging problem, particularly in large environments, as for example the devastated area after a disaster. Under such conditions standard methods for Simultaneous Localization And Mapping (SLAM) are difficult to apply due to possible misinterpretations of sensor data, leading to erroneous data association for loop closure. We consider the problem of multi-robot range-only SLAM for robot teams by solving the data association problem with wireless sensor nodes that we designed for this purpose. The memory of these nodes is utilized for the exchange of map data between multiple robots, facilitating loop-closures on jointly generated maps. We introduce RSLAM, which is a variant of FastSlam, extended for range-only measurements and the multi-robot case. Maps are generated from robot odometry and range estimates, which are computed from the RSSI (Received Signal Strength Indication). The proposed method has been extensively tested in USARSim, which serves as basis for the Virtual Robots competition at RoboCup, and by real-world experiments with a team of mobile robots. The presented results indicates that the approach is capable of building consistent maps in presence of real sensor noise, as well as to improve mapping results of multiple robots by data sharing.

• Alexander Kleiner, G. Steinbauer and F. Wotawa.
  Towards automated online diagnosis of robot navigation software.
  In Proc. of Int. Conf. on Simulation, Modeling and Programming for Autonomous Robots (SIMPAR), pp. 159-170. Springer 2008.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  Control software of autonomous mobile robots comprises a number of software modules that typically interact in a very complex way. Their proper interaction and the robustness of each single module strongly influences the safety during navigation in the field. Particularly in unstructured environments, unforeseen situations are likely to occur, causing erroneous behaviors of the robot. The proper handling of such situations requires an understanding of cause and effect within the complex interactions of the system. In this paper we present an approach which is able to automatically derive a model of the communication behavior within a component-orientated control software. The model can be used for online diagnosis in order to increase system robustness during runtime. We demonstrate model learning and system diagnosis on three different robot systems which were controlled by software modules communicating based on the widely used IPC (Inter Process Communication) standard. The demonstrated learning and diagnosis was carried out without any a priori knowledge about the systems.

• Christian Dornhege and Alexander Kleiner.
  Fully autonomous planning and obstacle negotiation on rough terrain using behavior maps.
  In Video Proceedings of the IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS 2007). San Diego, California 2007.
  

• Christian Dornhege and Alexander Kleiner.
  Behavior maps for online planning of obstacle negotiation and climbing on rough terrain.
  In Proceedings of the IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS 2007), pp. 3005-3011. 2007.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  To autonomously navigate on rough terrain is a challenging problem for mobile robots, requiring the ability to decide whether parts of the environment can be traversed or have to be bypassed, which is commonly known as Obstacle Negotiation (ON). In this paper, we introduce a planning framework that extends ON to the general case, where different types of terrain classes directly map to specific robot skills, such as climbing stairs and ramps. This extension is based on a new concept called behavior maps, which is utilized for the planning and execution of complex skills. Behavior maps are directly generated from elevation maps, i.e. two-dimensional grids storing in each cell the corresponding height of the terrain surface, and a set of skill descriptions. Results from extensive experiments are presented, showing that the method enables the robot to explore successfully rough terrain in real-time, while selecting the optimal trajectory in terms of costs for navigation and skill execution.

• Alexander Kleiner and R. Kümmerle.
  Genetic MRF model optimization for real-time victim detection in Search and Rescue.
  In Proceedings of the IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS 2007), pp. 3025-3030. 2007.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  One primary goal in rescue robotics is to deploy a team of robots for coordinated victim search after a disaster. This requires robots to perform subtasks, such as victim detection, in real-time. Human detection by computationally cheap techniques, such as color thresholding, turn out to produce a large number of false-positives. Markov Random Fields (MRFs) can be utilized to combine the local evidence of multiple weak classifiers in order to improve the detection rate. However, inference in MRFs is computational expensive. In this paper we present a novel approach for the genetic optimizing of the building process of MRF models. The genetic algorithm determines offline relevant neighborhood relations with respect to the data, which are then utilized for generating efficient MRF models from video streams during runtime. Experimental results clearly show that compared to a Support Vector Machine (SVM) based classifier, the optimized MRF models significantly reduce the false-positive rate. Furthermore, the optimized models turned out to be up to five times faster then the non-optimized ones at nearly the same detection rate.

• Alexander Kleiner and Christian Dornhege.
  Real-time Localization and Elevation Mapping within Urban Search and Rescue Scenarios.
  Journal of Field Robotics 24, pp. 723-745. 2007.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  Urban Search And Rescue (USAR) is a time critical task. Rescue teams have to explore a large terrain within a short amount of time in order to locate survivors after a disaster. One goal in Rescue Robotics is to have a team of heterogeneous robots that explore autonomously, or partially guided by an incident commander, the disaster area. Their task is to jointly create a map of the terrain and to register victim locations, which can further be utilized by human task forces for rescue. Basically, the robots have to solve autonomously in real-time the problem of Simultaneous Localization and Mapping (SLAM), consisting of a continuous state estimation problem and a discrete data association problem. Extraordinary circumstances after a real disaster make it very hard to apply common techniques. Many of these have been developed under strong assumptions, for example, they require polygonal structures, such as typically found in office-like environments. Furthermore, most techniques are not deployable in real-time. In this paper we propose real-time solutions for localization and mapping, which all have been extensively evaluated within the test arenas of the National Institute of Standards and Technology (NIST). We specifically deal with the problems of vision-based pose tracking on tracked vehicles, the building of globally consistent maps based on a network of RFID tags, and the building of elevation maps from readings of a tilted Laser Range Finder (LRF). Our results show that these methods lead under modest computational requirements to good results within the utilized testing arenas.

• S. Balakirsky, S. Carpin, Alexander Kleiner, M. Lewis, A. Visser, J. Wang and V.A. Ziparo.
  Towards heterogeneous robot teams for disaster mitigation: Results and Performance Metrics from Robocup Rescue.
  Journal of Field Robotics 24, pp. 943-967. 2007.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  Urban Search And Rescue is a growing area of robotic research. The RoboCup Federation has recognized this, and has created the new Virtual Robots competition to complement its existing physical robot and agent competitions. In order to successfully compete in this competition, teams need to field multi-robot solutions that cooperatively explore and map an environment while searching for victims. This paper presents the results of the first annual RoboCup Rescue Virtual competition. It provides details on the metrics used to judge the contestants as well as summaries of the algorithms used by the top four teams. This allows readers to compare and contrast these effective approaches. Furthermore, the simulation engine itself is examined and real-world validation results on the engine and algorithms are offered.

• V.A. Ziparo, Alexander Kleiner, L. Marchetti, A. Farinelli and and D. Nardi.
  Cooperative Exploration for USAR Robots with Indirect Communication.
  In Proceedings of the 6th IFAC Symposium on Intelligent Autonomous Vehicles (IAV 2007). 2007.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  To coordinate a team of robots for exploration is a challenging problem, particularly in unstructured areas, as for example post-disaster scenarios where direct communication is severely constrained. Furthermore, conventional methods of SLAM, e.g. those performing data association based on visual features, are doomed to fail due to bad visibility caused by smoke and fire. We use indirect communication (based on RFIDs), to share knowledge and use a gradient-like local search to direct robots towards interesting areas. To share a common frame of reference among robots we use a feature based SLAM approach (where features are RFIDs). The approach has been evaluated on a 3D simulation based on USARSim.

• Vittorio Ziparo, Alexander Kleiner, Bernhard Nebel and Daniele Nardi.
  RFID-Based Exploration for Large Robot Teams.
  In Proceedings of the IEEE International Conference on Robotics and Automation (ICRA 2007), pp. 4606-4613. Rome, Italy 2007.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  To coordinate a team of robots for exploration is a challenging problem, particularly in large areas as for example the devastated area after a disaster. This problem can generally be decomposed into task assignment and multi-robot path planning. In this paper, we address both problems jointly. This is possible because we reduce significantly the size of the search space by utilizing RFID tags as coordination points. The exploration approach consists of two parts: a stand-alone distributed local search and a global monitoring process which can be used to restart the local search in more convenient locations. Our results show that the local exploration works for large robot teams, particularly if there are limited computational resources. Experiments with the global approach showed that the number of conflicts can be reduced, and that the global coordination mechanism increases significantly the explored area.

• Alexander Kleiner, Christian Dornhege and Dali Sun.
  Mapping disaster areas jointly: RFID -Coordinated SLAM by Humans and Robots.
  In Proceedings of the IEEE International Workshop on Safety, Security and Rescue Robotics (SSRR 2007), pp. 1-6. 2007.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  We consider the problem of jointly performing SLAM by humans and robots in Urban Search And Rescue (USAR) scenarios. In this context, SLAM is a challenging task. First, places are hardly re-observable by vision techniques since visibility might be affected by smoke and fire. Second, loop-closure is cumbersome due to the fact that firemen will intentionally try to avoid performing loops when facing the reality of emergency response, e.g.USAR, while they are searching for victims. Furthermore, there might be places that are only accessible to robots, making it necessary to integrate humans and robots into one team for mapping the area after a disaster. In this paper, we introduce a method for jointly correcting individual trajectories of humans and robots by utilizing RFID technology for data association. Hereby the poses of humans and robots are tracked by a PDR (Pedestrian Dead Reckoning), and slippage sensitive odometry, respectively. We conducted extensive experiments with a team of humans, and a human-robot team within a semi-outdoor environment. Results from these experiments show that the introduced method allows to improve single trajectories based on the joint graph, even if they do not contain any loop.

• H. Kenn and Alexander Kleiner.
  Towards the Integration of Real-Time Real-World Data in Urban Search and Rescue Simulation.
  In MobileResponse, pp. 106-115. Springer 2007.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  The coordinated reaction to a large-scale disaster is a challenging research problem. The Robocup rescue simulation league addresses this research problem but is currently lacking an interface to real-world real-time data to test the validity of both simulation and simulated reactions. In this paper, we describe a wearable-computing-based real world interface to the Robocup Resuce simulation software and provide some updated results of preliminary evaluations.

• Alexander Kleiner and Dali Sun.
  Decentralized SLAM for Pedestrians without direct Communication.
  In Proceedings of the IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS 2007), pp. 1461-1466. 2007.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  We consider the problem of Decentralized Simultaneous Localization And Mapping (DSLAM) for pedestrians in the context of Urban Search And Rescue (USAR). In this context, DSLAM is a challenging task. First, data exchange fails due to cut off communication links. Second, loop-closure is cumbersome due to the fact that fireman will intentionally try to avoid performing loops, when facing the reality of emergency response, e.g. while they are searching for victims. In this paper, we introduce a solution to this problem based on the non-selfish sharing of information between pedestrians for loop-closure. We introduce a novel DSLAM method which is based on data exchange and association via RFID technology, not requiring any radio communication. The approach has been evaluated within both outdoor and semi-indoor environments. The presented results show that sharing information between single pedestrians allows to optimize globally their individual paths, even if they are not able to communicate directly.

• Alexander Kleiner, Johann Prediger and Bernhard Nebel.
  RFID Technology-based Exploration and SLAM for Search And Rescue.
  In Proceedings of the IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS 2006), pp. 4054-4059. Beijing, China 2006.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  Robot search and rescue is a time critical task, i.e. a large terrain has to be explored by multiple robots within a short amount of time. The efficiency of exploration depends mainly on the coordination between the robots and hence on the reliability of communication, which considerably suffers under the hostile conditions encountered after a disaster. Furthermore, rescue robots have to generate a map of the environment which has to be sufficiently accurate for reporting the locations of victims to human task forces. Basically, the robots have to solve autonomously in real-time the problem of Simultaneous Localization and Mapping (SLAM). This paper proposes a novel method for real-time exploration and SLAM based on RFID tags that are autonomously distributed in the environment. We utilized the algorithm of Lu and Milios [8] for calculating globally consistent maps from detected RFID tags. Furthermore we show how RFID tags can be used for coordinating the exploration of multiple robots. Results from experiments conducted in the simulation and on a robot show that our approach allows the computationally efficient construction of a map within harsh environments, and coordinated exploration of a team of robots.

• Alexander Kleiner, Christian Dornhege, Rainer Kuemmerle, Michael Ruhnke, Bastian Steder, Bernhard Nebel, Patrick Doherty, Mariusz Wzorek, Piotr Rudol, Gianpaolo Conte, S. Durante and D. Lundstrom.
  RoboCupRescue - Robot League Team RescueRobots Freiburg (Germany), Team Description Paper.
  In CDROM Proceedings of the International RoboCup Symposium '05. Bremen, Germany 2006.
  (Show abstract) (Hide abstract) (PDF)
  

  This paper describes the approach of the RescueRobots Freiburg team, which is a team of students from the University of Freiburg that originates from the former CS Freiburg team (RoboCupSoccer) and the ResQ Freiburg team (RoboCupRescue Simulation). Furthermore we introduce linkMAV, a micro aerial vehicle platform. Our approach covers RFID-based SLAM and exploration, autonomous detection of relevant 3D structures, visual odometry, and autonomous victim identification. Furthermore, we introduce a custom made 3D Laser Range Finder (LRF) and a novel mechanism for the active distribution of RFID tags.

• Alexander Kleiner and Vittorio Ziparo.
  RoboCupRescue - Simulation League Team RescueRobots Freiburg (Germany), Team Description Paper.
  In CDROM Proceedings of the International RoboCup Symposium '06. Bremen, Germany 2006.
  (PDF)
  

• Christian Dornhege and Alexander Kleiner.
  Visual Odometry for Tracked Vehicles.
  In Proceedings of the IEEE International Workshop on Safety, Security and Rescue Robotics (SSRR 2006). 2006.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  Localization and mapping on autonomous robots typically requires a good pose estimate, which is hard to acquire if the vehicle is tracked. In this paper we describe a solution to the pose estimation problem by utilizing a consumer-quality camera and an Inertial Measurement Unit (IMU). The basic idea is to continuously track salient features with the KLT feature tracker over multiple images taken by the camera and to extract from the tracked features image vectors resulting from the robot's motion. Each image vector is taken for a voting that best explains the robot's motion. Image vectors vote according to a previously trained ^\m2102tile coding^\m2112 classificator that assigns to each possible image vector a translation probability. Our results show that the proposed single camera solution leads to sufficiently accurate pose estimates of the tracked vehicle.

• Alexander Kleiner, N. Behrens and H. Kenn.
  Wearable computing meets multiagent systems: A real-world interface for the RoboCupRescue simulation platform.
  In First International Workshop on Agent Technology for Disaster Management at AAMAS06, pp. 116-123. AAMAS Press 2006.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  One big challenge in disaster response is to get an overview over the degree of damage and to provide this information, together with optimized plans for rescue missions, back to teams in the field. Collapsing infrastructure, limited visibility due to smoke and dust, and overloaded communication lines make it nearly impossible for rescue teams to report the total situation consistently. This problem can only be solved by efficiently integrating data of ^\m2102many^\m2112 observers into a single consistent view. A Global Positioning System (GPS) device in conjunction with a communication device, and sensors or simple input methods for reporting observations, offer a realistic chance to solve the data integration problem. We propose preliminary results from a wearable computing device, acquiring disaster relevant data, such as locations of victims and blockades, and show the data integration into the RoboCupRescue Simulation platform, which is a benchmark for MAS within the RoboCup competitions. We show exemplarily how the data can consistently be integrated and how rescue missions can be optimized by solutions developed on the RoboCupRescue simulation platform. The preliminary results indicate that nowadays wearable computing technology combined with MAS technology can serve as a powerful tool for Urban Search and Rescue (USAR).

• Alexander Kleiner, Michael Brenner, Tobias Braeuer, Christian Dornhege, Moritz Göbelbecker, Matthias Luber, Johann Prediger, Joerg Stueckler and Bernhard Nebel.
  Successful Search and Rescue in Simulated Disaster Areas.
  In Proceedings of the International RoboCup Symposium '05. Osaka, Japan 2005.
  (Show abstract) (Hide abstract) (PDF)
  

  RoboCupRescue Simulation is a large-scale multi-agent simulation of urban disasters where, in order to save lives and minimize damage, rescue teams must effectively cooperate despite sensing and communication limitations. This paper presents the comprehensive search and rescue approach of the ResQ Freiburg team, the winner in the RoboCupRescue Simulation league at RoboCup 2004. Specific contributions include the predictions of travel costs and civilian lifetime, the efficient coordination of an active disaster space exploration, as well as an any-time rescue sequence optimization based on a genetic algorithm. We compare the performances of our team and others in terms of their capability of extinguishing fires, freeing roads from debris, disaster space exploration, and civilian rescue. The evaluation is carried out with information extracted from simulation log files gathered during RoboCup 2004. Our results clearly explain the success of our team, and also confirm the scientific approaches proposed in this paper.

• Alexander Kleiner, Bastian Steder, Christian Dornhege, Daniel Hoefler, Daniel Meyer-Delius, Johann Prediger, Joerg Stueckler, Kolja Glogowski, Markus Thurner, Matthias Luber, Michael Schnell, Rainer Kuemmerle, Timothy Burk, Tobias Braeuer and Bernhard Nebel.
  RoboCupRescue - Robot League Team RescueRobots Freiburg (Germany), Team Description Paper.
  In CDROM Proceedings of the International RoboCup Symposium '05. Osaka, Japan 2005.
  (Show abstract) (Hide abstract) (PDF)
  

  This paper describes the approach of the RescueRobots Freiburg team. RescueRobots Freiburg is a team of students from the university of Freiburg, that originates from the former CS Freiburg team (RoboCupSoccer) and the ResQ Freiburg team (RoboCupRescue Simulation). Due to the high versatility of the RoboCupRescue competition we tackle the three arenas by a a twofold approach: On the one hand we want to introduce robust vehicles that can safely be teleoperated through rubble and building debris while constructing three-dimensional maps of the environment. On the other hand we want to introduce a team of autonomous robots that quickly explore a large terrain while building a two-dimensional map. This two solutions are particularly wellsuited for the red and yellow arena, respectively. Our solution for the orange arena will finally be decided between these two, depending on the capabilities of both approaches at the venue. In this paper, we introduce some preliminary results that we achieved so far from map building, localization, and autonomous victim identification. Furthermore we introduce a custom made 3D Laser Range Finder (LRF) and a novel mechanism for the active distribution of RFID tags. 1 Introduction RescueRobots Freiburg is a team of students from the university of Freiburg. The team originates from the former CS Freiburg team[6], which won three times the RoboCup world championship in the RoboCupSoccer F2000 league, and the ResQ Freiburg team[2], which won the last RoboCup world championship in the RoboCupRescue Simulation league. The team approach proposed in this paper is based on experiences gathered at RoboCup during the last six years. Due to the high versatility of the RoboCupRescue competition we tackle the three arenas by a twofold approach: On the one hand we want to introduce a vehicle that can safely be teleoperated through rubble and building debris while constructing threedimensional maps of the environment. On the other hand we want to introduce an autonomous team of robots that quickly explore a large terrain while building a twodimensional map. This two solutions are particularly well-suited for the red and yellow arena, respectively. Our solution for the orange arena will finally be decided between these two, depending on the capabilities of both approaches at the venue.

• Alexander Kleiner.
  Game AI: The shrinking gap between computer games and AI systems Ambient Intelligence.
  Ambient Intelligence:The evolution of technology, communication and cognition towards the future of human-computer interaction. 2005.
  (PDF)
  

• Timo Nuessle, Alexander Kleiner and Michael Brenner.
  Approaching Urban Disaster Reality: The ResQ Firesimulator.
  In Proceedings of the International RoboCup Symposium '04. Lisbon, Portugal 2004.
  (PDF)
  

• Alexander Kleiner, Michael Brenner, Tobias Braeuer, Christian Dornhege, Moritz Göbelbecker, Matthias Luber, Johann Prediger and Joerg Stueckler.
  ResQ Freiburg: Team Description and Evaluation, Team Description Paper, Rescue Simulation League.
  In CDROM Proceedings of the International RoboCup Symposium '04. Lisbon, Portugal 2004.
  (PDF)
  

• Erik Schulenburg, Thilo Weigel and Alexander Kleiner.
  Self-Localization in Dynamic Environments based on Laser and Vision Data.
  In Proceedings of the IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS'03), pp. 998-1004. Las Vegas, USA 2003.
  (PS.GZ) (PDF)
  

• Alexander Kleiner and Thorsten Buchheim.
  A Plugin-Based Architecture For Simulation In The F2000 League.
  In Proceedings of the International RoboCup Symposium '03. Padova, Italy 2003.
  (PS.GZ) (PDF)
  

• Alexander Kleiner, Markus Dietl and Bernhard Nebel.
  Towards a Life-Long Learning Soccer Agent.
  In Proceedings of the International RoboCup Symposium '02. Fukuoka, Japan 2002.
  (Show abstract) (Hide abstract) (PDF)
  

  One problem in robotic soccer (and in robotics in general) is to adapt skills and the overall behavior to a changing environment and to hardware improvements. We applied hierarchical reinforcement learning in an SMDP framework learning on all levels simultaneously. As our experiments show, learning simultaneously on the skill level and on the skill selection level is advantageous since it allows for a smooth adaption to a changing environment. Furthermore, the skills we trained turn also out to be quite competitive when run on the real robotic players of the players of our CS Freiburg team.

• Thilo Weigel, Jens-Steffen Gutmann, Markus Dietl, Alexander Kleiner and Bernhard Nebel.
  CS Freiburg: Coordinating Robots for Successful Soccer Playing.
  IEEE Transactions on Robotics and Automation 18 (5), pp. 685-699. 2002.
  (Show abstract) (Hide abstract) (PDF)
  

  Robotic soccer is a challenging research domain because many different research areas have to be addressed in order to create a successful team of robot players. This paper presents the CS Freiburg team, the winner in the middle size league at RoboCup 1998, 2000 and 2001. The paper focuses on multi-agent coordination for both perception and action. The contributions of this work are new methods for tracking ball and players observed by multiple robots, team coordination methods for strategic team formation and dynamic role assignment, a rich set of basic skills allowing to respond to large range of situations in an appropriate way, an action selection method based on behavior networks as well as a method to learn the skills and their selection. As demonstrated by evaluations of the different methods and by the success of the team, these methods permit the creation of a multi-robot group, which is able to play soccer successfully. In addition, the developed methods promise to advance the state of the art in the multi-robot field.

• Thilo Weigel, Alexander Kleiner, Florian Diesch, Markus Dietl, Jens-Steffen Gutmann, Bernhard Nebel, Patrick Stiegeler and Boris Szerbakowski.
  CS Freiburg 2001.
  In International RoboCup Symposium 2001. 2001.
  (Show abstract) (Hide abstract) (PDF)
  

  The CS Freiburg team has become F2000 champion the third time in the history of RoboCup. The success of our team can probably be attributed to its robust sensor interpretation and its team play. In this paper, we will focus on new developments in our vision system, in our path planner, and in the cooperation component.

Jana Koehler

• Jana Koehler and Jörg Hoffmann.
  On the Instantation of ADL Operators Involving Arbitrary First-Order Formulas.
  In Proceedings of the 14th Workshop on New Results in Planning, Scheduling and Design (PuK 2000) at ECAI 2000, pp. 74-82. Berlin, Germany 2000.
  (PS.GZ)
  

• Jana Koehler and Jörg Hoffmann.
  On Reasonable and Forced Goal Orderings and their Use in an Agenda-Driven Planning Algorithm.
  Journal of Artificial Intelligence Research 12, pp. 338-386. 2000.
  (PS.GZ)
  

• Jana Koehler and Jörg Hoffmann.
  Planning with Goal Agendas.
  In Proceedings des 13. Workshops Planen und Konfigurieren (PuK 1999) auf der 10. Tagung Expertensysteme (XPS-99). Würzburg, Germany 1999.
  (PS.GZ)
  

• Jörg Hoffmann and Jana Koehler.
  A new Method to Query and Index Sets.
  In Proceedings of the 16th International Joint Conference on Artificial Intelligence (IJCAI 1999). Stockholm, Sweden 1999.
  (PS.GZ) (extended technical report; PS.GZ)
  

• Hans Jürgen Ohlbach and Jana Koehler.
  Modal Logics, Description Logics and Arithmetic Reasoning.
  Artificial Intelligence 109 (1-2), pp. 1-31. 1999.
  (PS.GZ)
  

• Jana Koehler.
  Solving Complex Planning Tasks Through Extraction of Subproblems.
  In Proceedings of the 4th International Conference on Artificial Intelligence Planning Systems (AIPS-98). 1998.
  (PS.GZ)
  

• Jana Koehler.
  Planning under Resource Constraints.
  In Proceedings of the 13th European Conference on Artificial Intelligence (ECAI'98). 1998.
  (PS.GZ)
  

• Yannis Dimopoulos, Bernhard Nebel and Jana Koehler.
  Encoding planning problems in non-monotonic logic programs.
  In Proc. European Conference on Planning 1997 (ECP-97), pp. 169-181. Springer-Verlag 1997.
  (Show abstract) (Hide abstract) (PDF)
  

  In this paper we study a framework for encoding planning problems in logic programs with negation as failure. In contrast to other research work that focuses on the representional adequacy of nonmontonic logic programming as a language for describing theories of action and change, here we are concerned with more practical issues. Namely, we examine the effectiveness of an existing implementation of the stable models semantics called "Smodels" in solving a series of hard planning problems. We discuss representational issues and point out factors that can influence the performance of the method. It turns out that for careful and compact encodings, the performance of the method in a number of different domains, is comparable to that of planners like GRAPHPLAN and SATPLAN.

• Jana Koehler, Bernhard Nebel, Jörg Hoffmann and Yannis Dimopoulos.
  Extending Planning Graphs to an ADL Subset.
  In Proc. European Conference on Planning 1997 (ECP-97), pp. 273-285. Springer-Verlag 1997.
  (Show abstract) (Hide abstract) (PDF)
  

  We describe an extension of GRAPHPLAN to a subset of ADL that allows conditional and universally quantified effects in operators in such a way that almost all interesting properties of the original Graphplan algorithm are preserved.

• Bernhard Nebel, Yannis Dimopoulos and Jana Koehler.
  Ignoring Irrelevant Facts and Operators in Plan Generation.
  In Proc. European Conference on Planning 1997 (ECP-97), pp. 338-350. Springer-Verlag 1997.
  (Show abstract) (Hide abstract) (PDF)
  

  It is traditional wisdom that one should start from the goals when generating a plan in order to focus the plan generation process on potentially relevant actions. The GRAPHPLAN system, however, which is the most efficient planning system nowadays, builds a ``planning graph'' in a forward-chaining manner. Although this strategy seems to work well, it may possibly lead to problems if the planning task description contains irrelevant information. Although some irrelevant information can be filtered out by GRAPHPLAN, most cases of irrelevance are not noticed.

  In this paper, we analyze the effects arising from ``irrelevant'' information to planning task descriptions for different types of planners. Based on that, we propose a family of heuristics that select relevant information by minimizing the number of initial facts that are used when approximating a plan by backchaining from the goals ignoring any conflicts. These heuristics, although not solution-preserving, turn out to be very useful for guiding the planning process, as shown by applying the heuristics to a large number of examples from the literature.

• Hans Jürgen Ohlbach and Jana Koehler.
  Role Hierarchies and Number Restrictions.
  In Proc. Int. Workshop on Description Logics '97 (DL'97). 1997.
  (PS.GZ) (extended technical report; PS.GZ)
  

Sebastian Kupferschmid

• Sebastian Kupferschmid and Martin Wehrle.
  Abstractions and Pattern Databases: The Quest for Succinctness and Accuracy.
  In Parosh A. Abdulla and K. Rustan M. Leino (ed.), Proceedings of the 17th International Conference on Tools and Algorithms for the Construction and Analysis of Systems (TACAS 2011), pp. 276-290. Springer-Verlag 2011.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  Directed model checking is a well-established technique for detecting error states in concurrent systems efficiently. As error traces are important for debugging purposes, it is preferable to find as short error traces as possible. A wide spread method to find provably shortest error traces is to apply the A* search algorithm with distance heuristics that never overestimate the real error distance. An important class of such distance estimators is the class of pattern database heuristics, which are built on abstractions of the system under consideration. In this paper, we propose a systematic approach for the construction of pattern database heuristics. We formally define a concept to measure the accuracy of abstractions. Based on this technique, we address the challenge of finding abstractions that are succinct on the one hand, and accurate to produce informed pattern databases on the other hand. We evaluate our approach on large and complex industrial problems. The experiments show that the resulting distance heuristic impressively advances the state of the art.

• Martin Wehrle and Sebastian Kupferschmid.
  Context-Enhanced Directed Model Checking.
  In Jaco van de Pol and Michael Weber (eds.), Proceedings of the 17th International SPIN Workshop on Model Checking Software (SPIN 2010), pp. 88-105. Springer-Verlag 2010.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  Directed model checking is a well-established technique to efficiently tackle the state explosion problem when the aim is to find error states in concurrent systems. Although directed model checking has proved to be very successful in the past, additional search techniques provide much potential to efficiently handle larger and larger systems. In this work, we propose a novel technique for traversing the state space based on interference contexts. The basic idea is to preferably explore transitions that interfere with previously applied transitions, whereas other transitions are deferred accordingly. Our approach is orthogonal to the model checking process and can be applied to a wide range of search methods. We have implemented our method and empirically evaluated its potential on a range of non-trivial case studies. Compared to standard model checking techniques, we are able to detect subtle bugs with shorter error traces, consuming less memory and time.

• Martin Wehrle, Sebastian Kupferschmid and Andreas Podelski.
  Transition-based Directed Model Checking.
  In Stefan Kowalewski and Anna Philippou (eds.), Proceedings of the 15th International Conference on Tools and Algorithms for the Construction and Analysis of Systems (TACAS 2009), pp. 186-200. Springer-Verlag 2009.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  Directed model checking is a well-established technique that is tailored to fast detection of system states that violate a given safety property. This is achieved by influencing the order in which states are explored during the state space traversal. The order is typically determined by an abstract distance function that estimates a state's distance to a nearest error state. In this paper, we propose a general enhancement to directed model checking based on the evaluation of state transitions. We present a schema, parametrized by an abstract distance function, to evaluate transitions and propose a new method for the state space traversal. Our framework can be applied automatically to a wide range of abstract distance functions. The empirical evaluation impressively shows its practical potential. Apparently, the new method identifies a sweet spot in the trade-off between scalability (memory consumption) and short error traces.

• Thomas Keller and Sebastian Kupferschmid.
  Automatic Bidding for the Game of Skat.
  In Andreas R. Dengel, Karsten Berns, Thomas M. Breuel, Frank Bomarius and Thomas R. Roth-Berghofer (eds.), Proceedings of the 31st Annual German Conference on Artificial Intelligence (KI 2008), pp. 95-102. Springer-Verlag 2008.
  (Show abstract) (Hide abstract) (BIB) (PDF)
  

  In recent years, researchers started to study the game of Skat. The strength of existing Skat playing programs is definitely the card play phase. The bidding phase, however, was treated quite poorly so far. This is a severe drawback since bidding abilities influence the overall playing performance drastically. In this paper we present a powerful bidding engine which is based on a k-nearest neighbor algorithm.

• Martin Wehrle, Sebastian Kupferschmid and Andreas Podelski.
  Useless Actions are Useful.
  In Jussi Rintanen, Bernhard Nebel, J. Christopher Beck and Eric Hansen (eds.), Proceedings of the 18th International Conference on Automated Planning and Scheduling (ICAPS 2008), pp. 388-395. AAAI Press 2008.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  Planning as heuristic search is a powerful approach to solving domain independent planning problems. In recent years, various successful heuristics and planners like FF, LPG, Fast Downward or SGPlan have been proposed in this context. However, as heuristics only estimate the distance to goal states, a general problem of heuristic search is the existence of plateaus in the search space topology which can cause the search process to degenerate. Additional techniques like helpful actions or preferred operators that evaluate the "usefulness" of actions are often successful strategies to support the search in such situations.

  In this paper, we introduce a general method to evaluate the usefulness of actions. We propose a technique to enhance heuristic search by identifying "useless" actions that are not needed to find optimal plans. In contrast to helpful actions or preferred operators that are specific to the FF and Causal Graph heuristic, respectively, our method can be combined with arbitrary heuristics. We show that this technique often yields significant performance improvements.

• Sebastian Kupferschmid, Martin Wehrle, Bernhard Nebel and Andreas Podelski.
  Faster than Uppaal?
  In A. Gupta and S. Malik (eds.), Proceedings of the 20th International Conference on Computer Aided Verification (CAV 2008), pp. 552-555. Springer-Verlag 2008.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  It is probably very hard to develop a new model checker that is faster than Uppaal for verifying correct timed automata. In fact, our tool Mcta does not even try to compete with Uppaal in this (i.e., Uppaal's) arena. Instead, Mcta is geared towards analyzing incorrect specifications of timed automata. It returns (shorter) error traces faster.

• Sebastian Kupferschmid, Jörg Hoffmann and Kim G. Larsen.
  Fast Directed Model Checking via Russian Doll Abstraction.
  In C. R. Ramakrishnan and J. Rehof (eds.), Proceedings of the 14th International Conference on Tools and Algorithms for the Construction and Analysis of Systems (TACAS 2008), pp. 203-217. Springer-Verlag 2008.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  Directed model checking aims at speeding up the search for bugs in a system through the use of heuristic functions. Such a function maps states to integers, estimating the state's distance to the nearest error state. The search gives a preference to states with lower estimates. The key issue is how to generate good heuristic functions, i.e., functions that guide the search quickly to an error state. An arsenal of heuristic functions has been developed in recent years. Significant progress was made, but many problems still prove to be notoriously hard. In particular, a body of work describes heuristic functions for model checking timed automata in Uppaal, and tested them on a certain set of benchmarks. Into this arsenal we add another heuristic function. With previous heuristics, for the largest of the benchmarks it was only just possible to find some (unnecessarily long) error path. With the new heuristic, we can find provably shortest error paths for these benchmarks in a matter of seconds. The heuristic function is based on a kind of Russian Doll principle, where the heuristic for a given problem arises through using Uppaal itself for the complete exploration of a simplified instance of the same problem. The simplification consists in removing those parts from the problem that are distant from the error property. As our empirical results confirm, this simplification often preserves the characteristic structure leading to the error.

• Henning Dierks, Sebastian Kupferschmid and Kim G. Larsen.
  Automatic Abstraction Refinement for Timed Automata.
  In Jean-François Raskin and P. S. Thiagarajan (eds.), Proceedings of the 5th International Conference on Formal Modelling and Analysis of Timed Systems (FORMATS 2007), pp. 114-129. Springer-Verlag 2007.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  We present a fully automatic approach for counterexample guided abstraction refinement of real-time systems modelled in a subset of timed automata. Our approach is implemented in the Moby/RT tool environment, which is a CASE tool for embedded system specifications. Verification in Moby/RT is done by constructing abstractions of the semantics in terms of timed automata which are fed into the model checker Uppaal. Since the abstractions are over-approximations, absence of abstract counterexamples implies a valid result for the full model. Our new approach deals with the situation in which an abstract counterexample is found by Uppaal. The generated abstract counterexample is used to construct either a concrete counterexample for the full model or to identify a slightly refined abstraction in which the found spurious counterexample cannot occur anymore. Hence, the approach allows for a fully automatic abstraction refinement loop starting from the coarsest abstraction towards an abstraction for which a valid verification result is found. Nontrivial case studies demonstrate that this approach computes small abstractions fast without any user interaction.

• Sebastian Kupferschmid, Klaus Dräger, Jörg Hoffmann, Bernd Finkbeiner, Henning Dierks, Andreas Podelski and Gerd Behrmann.
  UPPAAL/DMC - Abstraction-based Heuristics for Directed Model Checking.
  In Orna Grumberg and Michael Huth (eds.), Proceedings of the 13th International Conference on Tools and Algorithms for the Construction and Analysis of Systems (TACAS 2007), pp. 679-682. Springer-Verlag 2007.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  Uppaal/DMC is an extension of Uppaal that provides generic heuristics for directed model checking. In this approach, the traversal of the state space is guided by a heuristic function which estimates the distance of a search state to the nearest error state. Our tool combines two recent approaches to design such estimation functions. Both are based on computing an abstraction of the system and using the error distance in this abstraction as the heuristic value. The abstractions, and thus the heuristic functions, are generated fully automatically and do not need any additional user input. Uppaal/DMC needs less time and memory to find shorter error paths than Uppaal's standard search methods.

• Jörg Hoffmann, Jan-Georg Smaus, Andrey Rybalchenko, Sebastian Kupferschmid and Andreas Podelski.
  Using Predicate Abstraction to Generate Heuristic Functions in Uppaal.
  In Stefan Edelkamp and Alessio Lomuscio (eds.), Proceedings of the 4th Workshop on Model Checking and Artificial Intelligence (MoChArt 2006), pp. 51-66. Springer-Verlag 2006.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  We focus on checking safety properties in networks of extended timed automata, with the well-known Uppaal system. We show how to use predicate abstraction, in the sense used in model checking, to generate search guidance, in the sense used in Artificial Intelligence (AI). This contributes another family of heuristic functions to the growing body of work on directed model checking. The overall methodology follows the pattern database approach from AI: the abstract state space is exhaustively built in a pre-process, and used as a lookup table during search. While typically pattern databases use rather primitive abstractions ignoring some of the relevant symbols, we use predicate abstraction, dividing the state space into equivalence classes with respect to a list of logical expressions (predicates). We empirically explore the behavior of the resulting family of heuristics, in a meaningful set of benchmarks. In particular, while several challenges remain open, we show that one can easily obtain heuristic functions that are competitive with the state-of-the-art in directed model checking.

• Sebastian Kupferschmid and Malte Helmert.
  A Skat Player Based on Monte Carlo Simulation.
  In H. Jaap van den Herik, Paolo Ciancarini and H. H. L. M. Donkers (eds.), Proceedings of the Fifth International Conference on Computer and Games (CG 2006), pp. 135-147. Springer-Verlag 2006.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  We apply Monte Carlo simulation and alpha-beta search to the card game of Skat, which is similar to Bridge, but different enough to require some new algorithmic ideas besides the techniques developed for Bridge. Our Skat-playing program integrates well-known techniques such as move ordering with two new search enhancements. Quasi-symmetry reduction generalizes symmetry reductions, popularized by Ginsberg's Partition Search algorithm, to search states which are "almost equivalent". Adversarial heuristics generalize ideas from single-agent search algorithms like A* to two-player games, leading to guaranteed lower and upper bounds for the score of a game position. Combining these techniques with state-of-the-art tree search algorithms, our program determines the game-theoretical value of a typical Skat hand (with perfect information) in 10 milliseconds.

• Sebastian Kupferschmid, Jörg Hoffmann, Henning Dierks and Gerd Behrmann.
  Adapting an AI Planning Heuristic for Directed Model Checking.
  In Antti Valmari (ed.), Proceedings of the 13th International SPIN Workshop on Model Checking Software (SPIN 2006), pp. 35-52. Springer-Verlag 2006.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  There is a growing body of work on directed model checking, which improves the falsification of safety properties by providing heuristic functions that can guide the search quickly towards short error paths. Techniques of this kind have also been made very successful in the area of AI Planning. Our main technical contribution is the adaptation of the most successful heuristic function from AI Planning to the model checking context, yielding a new heuristic for directed model checking. The heuristic is based on solving an abstracted problem in every search state. We adapt the abstraction and its solution to networks of communicating automata annotated with (constraints and effects on) integer variables. Since our ultimate goal in this research is to also take into account clock variables, as used in timed automata, our techniques are implemented inside Uppaal. We run experiments in some toy benchmarks for timed automata, and in two timed automata case studies originating from an industrial project. Compared to both blind search and some previously proposed heuristic functions, we consistently obtain significant, sometimes dramatic, search space reductions, resulting in likewise strong reductions of runtime and memory requirements.

• Jörg Hoffmann and Sebastian Kupferschmid.
  A Covering Problem for Hypercubes.
  In Leslie Pack Kaelbling and Alessandro Saffiotti (eds.), Poster Proceedings of the 19th International Joint Conference on Artificial Intelligence (IJCAI 2005), pp. 1523-1524. 2005.
  (Show abstract) (Hide abstract) (PDF) (PS.GZ) (BIB)
  

  We introduce a new NP-complete problem asking if a "query" hypercube is (not) covered by a set of other "evidence" hypercubes. This comes down to a form of constraint reasoning asking for the satisfiability of a CNF formula where the logical atoms are inequalities over single variables, with possibly infinite variable domains. We empirically investigate the location of the phase transition regions in two random distributions of problem instances. We introduce a solution method that iteratively constructs a representation of the non-covered part of the query cube. In particular, the method is not based on backtracking. Our experiments show that the method is, in a significant range of instances, superior to the backtracking method that results from translation to SAT, and application of a state-of-the-art DP-based SAT solver.

Christian Köhler

• Christian Köhler, Artur Ottlik, Hans-Hellmut Nagel and Bernhard Nebel.
  Qualitative Reasoning Feeding Back into Quantitative Model-Based Tracking.
  In Proceedings of the 16th European Conference on Artificial Intelligence (ECAI 2004), pp. 1041-1042. IOS Press 2004.
  (Show abstract) (Hide abstract) (PDF) (technical report; PDF)
  

  Tracking vehicles in image sequences of innercity road traffic scenes still constitutes a challenging task. Even if a-priori knowledge about the 3D shape of vehicles, of background structure and vehicle motion is provided, (partial) occlusion and dense vehicle queues easily can cause initialization and tracking failures. Improving the tracking approach requires numerous and time-consuming experiments. Yet, these difficulties can be eased considerably by endowing the system with a part of the qualitative knowledge, that a human observer uses in order to judge the results. In the case reported here, a system for qualitative reasoning has been coupled with a quantitative model-based tracking system in order to explore the feedback from qualitative reasoning into the geometric tracking subsystem.

• Christian Köhler.
  Selecting Ghosts and Queues from a Car Trackers Output using a Spatio-Temporal Query Language.
  In IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR 2004), pp. 619-624. Washington, D.C., USA 2004.
  (PDF) (PS.GZ)
  

• Christian Köhler.
  The Occlusion Calculus.
  In Proceedings Workshop on Cognitive Vision. Zürich, Switzerland 2002.
  (PS.GZ) (PDF)
  

Felix Lindner

• Felix Lindner, Robert Mattmüller and Bernhard Nebel.
  Evaluation of the Moral Permissibility of Action Plans.
  Artificial Intelligence 287. 2020.
  (Show abstract) (Hide abstract) (PDF)
  

  Research in classical planning so far has been mainly concerned with generating a satisficing or an optimal plan. However, if such systems are used to make decisions that are relevant to humans, one should also consider the ethical consequences generated plans can have. Traditionally, ethical principles are formulated in an action-based manner, allowing to judge the execution of one action. We show how such a judgment can be generalized to plans. Further, we study the computational complexity of making ethical judgment about plans.

• Felix Lindner, Barbara Kuhnert, Laura Wächter and Katrin Möllney.
  Perception of Creative Responses to Moral Dilemmas by a Conversational Robot.
  In Proc. ICSR 2019. 2019.
  (PDF)
  

• Felix Lindner and Katrin Möllney.
  Extracting Reasons for Moral Judgments under Various Ethical Principles.
  In Proceedings of KI 2019. 2019.
  (PDF)
  

• Hanna Stellmach and Felix Lindner.
  Perception of an Uncertain Ethical Reasoning Robot.
  Journal of Interactive Media 18(1). 2019.
  

• Felix Lindner, Robert Mattmüller and Bernhard Nebel.
  Moral Permissibility of Action Plans.
  In Proceedings of the Thirty-Third AAAI Conference on Artificial Intelligence (AAAI-19). 2019.
  (Show abstract) (Hide abstract) (PDF)
  

  Research in classical planning so far was mainly concerned with generating a satisficing or an optimal plan. However, if such systems are used to make decisions that are relevant to humans, one should also consider the ethical consequences generated plans can have. We address this challenge by analyzing in how far it is possible to generalize existing approaches of machine ethics to automatic planning systems. Traditionally, ethical principles are formulated in an action-based manner, allowing to judge the execution of one action. We show how such a judgment can be generalized to plans. Further, we study the computational complexity of making ethical judgment about plans.

• Laura Wächter and Felix Lindner.
  An Explorative Comparison of Blame Attributions to Companion Robots Across Various Moral Dilemmas.
  In Proceedings of The 6th International Conference on Human-Agent Interaction (HAI 2018). 2018.
  

• Barbara Kuhnert, Felix Lindner, Martin Mose Bentzen and Marco Ragni.
  Causal Structure of Moral Dilemmas Predicts Perceived Difficulty of Making a Decision.
  In Proceedings of the KogWis 2018 (Extended Abstract). 2018.
  

• Martin Mose Bentzen, Felix Lindner, Louise Dennis and Michael Fisher.
  Moral Permissibility of Actions in Smart Home Systems.
  In Proceedings of the FLoC 2018 Workshop on Robots, Morality, and Trust through the Verification Lens (Extended Abstract). 2018.
  

• Hanna Stellmach and Felix Lindner.
  Perception of an Uncertain Ethical Reasoning Robot: A Pilot Study.
  In Proceedings of Mensch und Computer 2018. 2018.
  (Show abstract) (Hide abstract)
  

  The study investigates the effect of uncertainty expressed by a robot facing a moral dilemma. Participants (N = 80) were shown a video of a robot explaining a moral dilemma and the decision it makes. The robot either expressed certainty or uncertainty about its decision. Participants rated how much blame the robot deserves for its action, the moral wrongness of the action, and their impression of the robot in terms of four scale dimensions measuring social perception. The results suggest that participants that were not familiar with the moral dilemma assign more blame to the robot for the same action when it expresses uncertainty, while expressed uncertainty has less effect on moral wrongness judgments. There was no significant effect of expressed uncertainty on participants’ impression of the robot. We discuss implications of this result for the design of social robots.

• Felix Lindner and Martin Mose Bentzen.
  A Formalization of Kant's Second Formulation of the Categorical Imperative.
  In Proceedings of the 14th International Conference on Deontic Logic and Normative Systems (DEON). 2018.
  (Show abstract) (Hide abstract)
  

  We present a formalization and computational implementation of the second formulation of Kant's categorical imperative. This ethical principle requires an agent to never treat someone merely as a means but always also as an end. Here we interpret this principle in terms of how persons are causally affected by actions. We introduce Kantian causal agency models in which moral patients, actions, goals, and causal influence are represented, and we show how to formalize several readings of Kant's categorical imperative that correspond to Kant's concept of strict and wide duties towards oneself and others. Stricter versions handle cases where an action directly causally affects oneself or others, whereas the wide version maximizes the number of persons being treated as an end. We discuss limitations of our formalization by pointing to one of Kant's cases that the machinery cannot handle in a satisfying way.

• Felix Lindner, Robert Mattmüller and Bernhard Nebel.
  Moral Permissibility of Action Plans.
  In Proceedings of the ICAPS Workshop on EXplainable AI Planning (XAIP). 2018.
  (Show abstract) (Hide abstract) (PDF)
  

  Research in classical planning so far was mainly concerned with generating a satisficing or an optimal plan. However, if such systems are used to make decisions that are relevant to humans, one should also consider the ethical consequences that generated plans can have. We address this challenge by analyzing in how far it is possible to generalize existing approaches of machine ethics to automatic planning systems. Traditionally, ethical principles are formulated in an action-based manner, allowing to judge the execution of one action. We show how such a judgment can be generalized to plans. Further, we study the complexity of making ethical judgment about plans.

• Glenda Hannibal and Felix Lindner.
  Transdisciplinary Reflections on Social Robotics in Academia and Beyond.
  In Proceedings of Robo-Philosophy 2018. 2018.
  

• Felix Lindner, Martin Mose Bentzen and Bernhard Nebel.
  The HERA Approach to Morally Competent Robots.
  In Proceedings of the 2017 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS 2017). 2017.
  (Show abstract) (Hide abstract) (PDF)
  

  To address the requirement for autonomous moral decision making, we introduce a software library for modeling hybrid ethical reasoning agents (short: HERA). The goal of the HERA project is to provide theoretically well-founded and practically usable logic-based machine ethics tools for implementation in robots. The novelty is that HERA implements multiple ethical principles like utilitarianism, the principle of double effect, and a Pareto-inspired principle. These principles can be used to automatically assess moral situations represented in a format we call causal agency models. We discuss how to model moral situations using our approach, and how it can cope with uncertainty about moral values. Finally, we briefly outline the architecture of our robot IMMANUEL, which implements HERA and is able to explain ethical decisions to humans.

• Barbara Kuhnert, Marco Ragni and Felix Lindner.
  The Gap between Human's Attitute towards Robots in General and Human's Expectation of an Ideal Everyday Life Robot.
  In Proceedings of the 2017 IEEE International Symposium on Robot and Human Interactive Communication (RO-MAN 2017). 2017.
  (PDF)
  

• Felix Lindner, Laura Wächter and Martin Mose Bentzen.
  Discussions About Lying with an Ethical Reasoning Robot.
  In Proceedings of the 2017 IEEE International Symposium on Robot and Human Interactive Communication (RO-MAN 2017). 2017.
  (PDF)
  

• Felix Lindner and Carola Eschenbach.
  An Affordance-Based Conceptual Framework for Spatial Behavior of Social Robots.
  In Raul Hakli and Johanna Seibt (eds.), Sociality and Normativity for Robots --- Philosophical Inquiries into Human-Robot Interactions. Springer International Publishing 2017.
  

• Barbara Kuhnert, Felix Lindner, Martin Mose Bentzen and Marco Ragni.
  Perceived Difficulty of Moral Dilemmas Depends on Their Causal Structure: A Formal Model and Preliminary Results.
  In Proceedings of the 39th Annual Meeting of the Cognitive Science Society CogSci 2017. 2017.
  (PDF)
  

• Felix Lindner and Martin Mose Bentzen.
  The Hybrid Ethical Reasoning Agent IMMANUEL.
  In Proceedings of the 2017 Conference on Human-Robot Interaction (HRI2017), Late-Breaking Report. 2017.
  (Show abstract) (Hide abstract) (PDF)
  

  We introduce a novel software library that supports the implementation of hybrid ethical reasoning agents (HERA). The objective is to make moral principles available to robot programming. At its current stage, HERA can assess the moral permissibility of actions according to the principle of double effect, utilitarianism, and the do-no-harm principle. We present the prototype robot IMMANUEL based on HERA. The robot will be used to conduct research on joint moral reasoning in human-robot interaction.

• Felix Lindner.
  How To Count Multiple Personal-Space Intrusions in Social Robot Navigation.
  In Proceedings of the Robo-Philosophy Conference 2016. 2016.
  (Show abstract) (Hide abstract)
  

  One aspect of social robot navigation is to avoid personal space intrusions. Computationally, this can be achieved by introducing social costs into a robot's path planner's objective function. This article tackles the normative question of how robots should aggregate social costs incurred by multiple personal-space intrusions. Of particular interest is the question whether numbers should count, i.e., whether a robot ought to intrude into one person's personal space in order to aboid intruding into multiple personal spaces. This work proposes four different modes of aggregation of the costs of intrusions into personal space, discusses some of the philosophical arguments, and presents results from a pilot study.

• Felix Lindner.
  A Model of a Robot's Will Based on Higher-Order Desires.
  In Proceedings of the IEEE International Symposium on Robot and Human Interactive Communication (RO-MAN 2016). 2016.
  (Show abstract) (Hide abstract)
  

  Autonomous robots implement decision making capacities on several layers of abstraction. Put in terms of desires, decision making evaluates desires to eventually commit to some most rational one. Drawing on the philosophical literature on volition and agency, this work introduces a conceptual model that enables robots to reason about which desires they want to want to realize, i.e., higher-order desires. As a result, six jointly exhaustive and pairwise disjoint types of choices are defined. Technical evaluation shows how to add a robot's will to its rational decision-making capacity. This guarantees that informed choices are possible even in cases rational decision making alone is indecisive. Further applications to modeling personality traits for human-robot interaction are discussed.

• Felix Lindner.
  A Social Robot's Knowledge About Territories in Public Space.
  In Proceedings of the 3rd Workshop on Public Space Human-Robot Interaction (PubRob 2016). 2016.
  (Show abstract) (Hide abstract)
  

  Human territoriality is a microsocial phenomenon that displays the strong interrelations between action and space. Actions are spatially located, and because spaces that afford certain activities are rare, humans claim portions of space and restrict space access to particular agents. Thereby, they create territories. In public spaces, territories are usually short-term yet bravely defended: E.g., the table in a restaurant, the place in a queue at the checkout, the seat in the train. This abstract sketches a formal theory of territory. Its aim is to enable social robots to consider existing territories during decision making and planning both in order to avoid intrusions of others' territories and to claim territories for their own benefit.

Johannes Löhr

• Johannes Löhr, Martin Wehrle, Maria Fox and Bernhard Nebel.
  Symbolic Domain Predictive Control.
  In Proceedings of the 28th AAAI Conference on Artificial Intelligence (AAAI 2014), pp. 2315-2321. AAAI Press 2014.
  (Show abstract) (Hide abstract) (Online; DOI)
  

  Planning-based methods to guide switched hybrid systems from an initial state into a desired goal region opens an interesting field for control. The idea of the Domain Predictive Control (DPC) approach is to generate input signals affecting both the numerical states and the modes of the system by stringing together atomic actions to a logically consistent plan. However, the existing DPC approach is restricted in the sense that a discrete and pre-defined input signal is required for each action. In this paper, we extend the approach to deal with symbolic states. This allows for the propagation of reachable regions of the state space emerging from actions with inputs that can be arbitrarily chosen within specified input bounds. This symbolic extension enables the applicability of DPC to systems with bounded inputs sets and increases its robustness due to the implicitly reduced search space. Moreover, precise numeric goal states instead of goal regions become reachable.

• Johannes Löhr, Johannes Aldinger, Stefan Winkler and Georg Willich.
  Automated Planning for Earth Observation Spacecraft under Attitude Dynamical Constraints.
  In Jahrbuch der Deutschen Gesellschaft für Luft- und Raumfahrt (DGLR2013). 2013.
  (Show abstract) (Hide abstract) (PDF)
  

  Agile Earth observation missions continuously require a large amount of planning during the spacecraft's observations. Beside priorities of the observation sites, especially the agility constraints of the satellite are important to be taken into account during the planning process. This is due to the body-fixed instrument's line of sight, requiring the whole satellite to point to the observation sites while scanning. Scanning a sequence of observation sites leads to complex slew maneuvers which must not exceed the satellite's actuator capacities, attitude constraints or maximum angular rates. Additionally, the regions of interest may change over time, making it necessary to adapt and optimize the observation sequence continuously. An automated process is required to efficiently handle this task. We present a planning algorithm to sequence an arbitrarily distributed set of observation patches to a feasible observation plan, considering priority criteria of the observation sites and agility constraints of the satellite.

• Johannes Aldinger and Johannes Löhr.
  Planning for Agile Earth Observation Satellites.
  In Proceedings of the ICAPS-2013 Workshop on Planning in Continuous Domains (PCD), pp. 9-17. 2013.
  (Show abstract) (Hide abstract) (PDF)
  

  Agile Earth observation satellites are satellites orbiting Earth with the purpose to gather information of the Earth's surface by slewing the satellite toward regions of interest. Constraints arise not only from dynamical and kinematic aspects of the satellite and its sensors. Regions of interest change over time and bad weather can conceal important observation targets. This results in a constant need to replan the satellite's tasks and raises the desire to automatize this planning process. We consider the Earth observation problem with the help of the module extension of the numerical planning system Temporal Fast Downward. Complex satellite slew maneuvers are calculated within modules, while the planner selects and schedules the regions to be scanned. First results encourage deeper research in this area so that forthcoming satellite space missions can draw on automated planning to improve the performance of agile Earth observation tasks.

• Johannes Löhr, Patrick Eyerich, Stefan Winkler and Bernhard Nebel.
  Domain Predictive Control Under Uncertain Numerical State Information.
  In Proceedings of the 23rd International Conference on Automated Planning and Scheduling (ICAPS13). 2013.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  In planning, hybrid system states consisting of logical and numerical variables are usually assumed to be completely known. In particular, for numerical state variables full knowledge of their exact values is assumed. However, in real world applications states are results of noisy measurements and imperfect actuators. Therefore, a planned sequence of state transitions might fail to lead a hybrid system to the desired goal. We show how to propagate and reason about uncertain state information directly in the planning process, enabling hybrid systems to find plans that satisfy numerical goals with predefined confidence.

• Johannes Löhr, Bernhard Nebel and Stefan Winkler.
  Planning Based Autonomous Lander Control.
  In Proceedings of the Astrodynamics Specialist Conference (AIAA/AAS 2012). 2012.
  (Show abstract) (Hide abstract) (Online; DOI)
  

  Safe landing of spacecraft on extraterrestrial surfaces implies a number of challenges. The main issue is to precisely initiate coasting, braking and landing maneuvers to safely land at a desired landing zone. Meanwhile, the increasing information level about the landing environment has to be processed and the landing trajectory eventually adapted in order to avoid hazardous situations. In this paper these time critical tasks are performed by Domain Predictive Control. It has been developed to guide dynamic systems into desired goal states by flexibly reordering atomic actions using planning algorithms from artificial intelligence. Here, the method is applied to autonomously adapt control commands and associated landing trajectories with respect to the changing environmental knowledge. Simulation results show the feasibility of this new approach and reveal issues which should be subject to future research.

• Johannes Löhr, Patrick Eyerich, Thomas Keller and Bernhard Nebel.
  A Planning Based Framework for Controlling Hybrid Systems.
  In Proceedings of the 22nd International Conference on Automated Planning and Scheduling (ICAPS 2012). 2012.
  (Show abstract) (Hide abstract) (PDF)
  

  The control of dynamic systems, which aims to minimize the deviation of state variables from reference values in a contin- uous state space, is a central domain of cybernetics and con- trol theory. The objective of action planning is to find feasible state trajectories in a discrete state space from an initial state to a state satisfying the goal conditions, which in principle ad- dresses the same issue on a more abstract level. We combine these approaches to switch between dynamic system charac- teristics on the fly, and to generate control input sequences that affect both discrete and continuous state variables. Our approach (called Domain Predictive Control) is applicable to hybrid systems with linear dynamics and discretizable inputs.

• Johannes Löhr and Stefan Winkler.
  Comparison of Periodic System Lifting Techniques for Robust Stability Analysis of Magnetic Spacecraft Attitude Control Systems.
  In Proceedings of the Guidance Navigation and Control Conference (AIAA/GNC 2011). 2011.
  (Show abstract) (Hide abstract)
  

  Magnetic attitude control is a common method for low Earth orbit spacecraft. Verification of nominal (yes/no), and even more important, robust attitude stability of such systems is of significant importance for any real mission. Considering nadir-oriented operation, the linearized closed-loop attitude dynamics equations lead to a linear time periodic system. While nominal stability can be obtained from Floquet-Theory, robust stability analysis via standard μ-Analysis requires lifting procedures to convert the linear time periodic into a linear time invariant system. Two of those lifting procedures are compared in this paper: (1) "Fast Discretization", which is based on a "sample and hold" view on the periodic state space matrices, and (2) "Frequency Lifting" based on Fourier series expansion. Both methods are applied to a theoretic linear time periodic example from literature and magnetic spacecraft attitude control. The comparison focus' on applicability for real satellite missions.

Robert Mattmüller

• Julian von Tschammer, Robert Mattmüller and David Speck.
  Loopless Top-k Planning.
  In Proceedings of the 32nd International Conference on Automated Planning and Scheduling (ICAPS 2022). 2022.
  (Show abstract) (Hide abstract) (PDF)
  

  In top-k planning, the objective is to determine a set of k cheapest plans that provide several good alternatives to choose from. Such a solution set often contains plans that visit at least one state more than once. Depending on the application, plans with such loops are of little importance because they are dominated by a loopless representative and can prevent more meaningful plans from being found.

  In this paper, we motivate and introduce loopless top-k planning. We show how to enhance the state-of-the-art symbolic top-k planner, symK, to obtain an efficient, sound and complete algorithm for loopless top-k planning. An empirical evaluation shows that our proposed approach has a higher k-coverage than a generate-and-test approach that uses an ordinary top-k planner, which we show to be incomplete in the presence of zero-cost loops.

• David Speck, David Borukhson, Robert Mattmüller and Bernhard Nebel.
  On the Compilability and Expressive Power of State-Dependent Action Costs.
  In Proceedings of the 31st International Conference on Automated Planning and Scheduling (ICAPS 2021), pp. 358-366. 2021.
  (Show abstract) (Hide abstract) (PDF)
  

  While state-dependent action costs are practically relevant and have been studied before, it is still unclear if they are an essential feature of planning tasks. In this paper, we study to what extent state-dependent action costs are an essential feature by analyzing under which circumstances they can be compiled away. We give a comprehensive classification for combinations of action cost functions and possible cost measures for the compilations.

  Our theoretical results show that if both task sizes and plan lengths are to be preserved polynomially, then the boundary between compilability and non-compilability lies between FP and FPSPACE computable action cost functions (under a mild assumption on the polynomial hierarchy). Preserving task sizes polynomially and plan lengths linearly at the same time is impossible.

• David Speck, André Biedenkapp, Frank Hutter, Robert Mattmüller and Marius Lindauer.
  Learning Heuristic Selection with Dynamic Algorithm Configuration.
  In Proceedings of the 31st International Conference on Automated Planning and Scheduling (ICAPS 2021), pp. 597-605. 2021.
  (Show abstract) (Hide abstract) (PDF)
  

  A key challenge in satisficing planning is to use multiple heuristics within one heuristic search. An aggregation of multiple heuristic estimates, for example by taking the maximum, has the disadvantage that bad estimates of a single heuristic can negatively affect the whole search. Since the performance of a heuristic varies from instance to instance, approaches such as algorithm selection can be successfully applied. In addition, alternating between multiple heuristics during the search makes it possible to use all heuristics equally and improve performance. However, all these approaches ignore the internal search dynamics of a planning system, which can help to select the most useful heuristics for the current expansion step. We show that dynamic algorithm configuration can be used for dynamic heuristic selection which takes into account the internal search dynamics of a planning system. Furthermore, we prove that this approach generalizes over existing approaches and that it can exponentially improve the performance of the heuristic search. To learn dynamic heuristic selection, we propose an approach based on reinforcement learning and show empirically that domain-wise learned policies, which take the internal search dynamics of a planning system into account, can exceed existing approaches.

• Thorsten Engesser, Robert Mattmüller, Bernhard Nebel and Michael Thielscher.
  Game description language and dynamic epistemic logic compared.
  Artificial Intelligence 292. 2021.
  (Show abstract) (Hide abstract) (Online; DOI)
  

  Several different frameworks have been proposed to model and reason about knowledge in dynamic multi-agent settings, among them the logic-programming-based game description language GDL-III and dynamic epistemic logic (DEL). GDL-III and DEL have complementary strengths and weaknesses in terms of ease of modeling and simplicity of semantics. In this paper, we formally study the expressiveness of GDL-III vs. DEL. We clarify the commonalities and differences between those languages, demonstrate how to bridge the differences where possible, and identify large fragments of GDL-III and DEL that are equivalent in the sense that they can be used to encode games or planning tasks that admit the same legal action sequences. We prove the latter by providing translations between those fragments of GDL-III and DEL.

• Patrick Caspari, Robert Mattmüller and Tim Schulte.
  A Framework to Prove Strong Privacy in Multi-Agent Planning.
  In Proceedings of the 6th Workshop on Distributed and Multi-Agent Planning (DMAP 2020), pp. 32-39. 2020.
  (PDF)
  

• David Speck, André Biedenkapp, Frank Hutter, Robert Mattmüller and Marius Lindauer.
  Learning Heuristic Selection with Dynamic Algorithm Configuration.
  In Proceedings of the Workshop on Bridging the Gap Between AI Planning and Reinforcement Learning (PRL 2020), p. 61–69. 2020.
  Superseded by the ICAPS 2021 paper by the same authors.
  (Show abstract) (Hide abstract) (PDF)
  

  A key challenge in satisficing planning is to use multiple heuristics within one heuristic search. An aggregation of multiple heuristic estimates, for example by taking the maximum, has the disadvantage that bad estimates of a single heuristic can negatively affect the whole search. Since the performance of a heuristic varies from instance to instance, approaches such as algorithm selection can be successfully applied. In addition, alternating between multiple heuristics during the search makes it possible to use all heuristics equally and improve performance. However, all these approaches ignore the internal search dynamics of a planning system, which can help to select the most helpful heuristics for the current expansion step. We show that dynamic algorithm configuration can be used for dynamic heuristic selection which takes into account the internal search dynamics of a planning system. Furthermore, we prove that this approach generalizes over existing approaches and that it can exponentially improve the performance of the heuristic search. To learn dynamic heuristic selection, we propose an approach based on reinforcement learning and show empirically that domain-wise learned policies, which take the internal search dynamics of a planning system into account, can exceed existing approaches in terms of coverage.

• Dominik Drexler, David Speck and Robert Mattmüller.
  Subset-Saturated Transition Cost Partitioning for Optimal Classical Planning.
  In Proceedings of the 12th Workshop on Heuristics and Search for Domain-Independent Planning (HSDIP 2020), p. 23–31. 2020.
  Superseded by the ICAPS 2021 paper "Subset-Saturated Transition Cost Partitioning".
  (Show abstract) (Hide abstract) (PDF)
  

  Cost partitioning admissibly combines the information from multiple heuristics for state-space search. We use a greedy method called saturated cost partitioning that considers the heuristics in sequence and assigns the minimal fraction of the remaining costs that it needs to preserve the heuristic estimates. In this work, we address the problem of using more expressive transition cost functions with saturated cost partitioning to obtain stronger heuristics. Our contribution is subset-saturated transition cost partitioning that combines the concepts of using transition cost functions and prioritizing states that look more important during the search. Our empirical evaluation shows that this approach still causes too much computational overhead but leads to more informed heuristics.

• Thorsten Engesser, Robert Mattmüller, Bernhard Nebel and Felicitas Ritter.
  Token-based Execution Semantics for Multi-Agent Epistemic Planning.
  In Proceedings of the 17th International Conference on Principles of Knowledge Representation and Reasoning (KR-2020), pp. 351-360. 2020.
  (Show abstract) (Hide abstract) (Online; PDF)
  

  Epistemic planning has been employed as a means to achieve implicit coordination in cooperative multi-agent systems where world knowledge is distributed between the agents, and agents plan and act individually. However, recent work has shown that even if all agents act with respect to plans that they consider optimal from their own subjective perspective, infinite executions can occur. In this paper, we analyze the idea of using a single token that can be passed around between the agents and which is used as a prerequisite for acting. We show that introducing such a token to any planning task will prevent the existence of infinite executions. We furthermore analyze the conditions under which solutions to a planning task are preserved under our tokenization.

• Felix Lindner, Robert Mattmüller and Bernhard Nebel.
  Evaluation of the Moral Permissibility of Action Plans.
  Artificial Intelligence 287. 2020.
  (Show abstract) (Hide abstract) (PDF)
  

  Research in classical planning so far has been mainly concerned with generating a satisficing or an optimal plan. However, if such systems are used to make decisions that are relevant to humans, one should also consider the ethical consequences generated plans can have. Traditionally, ethical principles are formulated in an action-based manner, allowing to judge the execution of one action. We show how such a judgment can be generalized to plans. Further, we study the computational complexity of making ethical judgment about plans.

• David Speck, Florian Geißer and Robert Mattmüller.
  When Perfect is not Good Enough: On the Search Behaviour of Symbolic Heuristic Search.
  In Proceedings of the 30th International Conference on Automated Planning and Scheduling (ICAPS 2020), pp. 263-271. 2020.
  (Show abstract) (Hide abstract) (PDF)
  

  Symbolic search has proven to be a competitive approach to cost-optimal planning, as it compactly represents sets of states by symbolic data structures. While heuristics for symbolic search exist, symbolic bidirectional blind search empirically outperforms its heuristic counterpart and is therefore the dominant search strategy. This prompts the question of why heuristics do not seem to pay off in symbolic search. As a first step in answering this question, we investigate the search behaviour of symbolic heuristic search by means of BDDA*. Previous work identified the partitioning of state sets according to their heuristic values as the main bottleneck. We theoretically and empirically evaluate the search behaviour of BDDA* and reveal another fundamental problem: we prove that the use of a heuristic does not always improve the search performance of BDDA*. In general, even the perfect heuristic can exponentially deteriorate search performance.

• David Speck, Robert Mattmüller and Bernhard Nebel.
  Symbolic Top-k Planning.
  In Proceedings of the 34th AAAI Conference on Artificial Intelligence (AAAI 2020), pp. 9967-9974. 2020.
  (Show abstract) (Hide abstract) (PDF)
  

  The objective of top-k planning is to determine a set of k different plans with lowest cost for a given planning task. In practice, such a set of best plans can be preferred to a single best plan generated by ordinary optimal planners, as it allows the user to choose between different alternatives and thus take into account preferences that may be difficult to model. In this paper we show that, in general, the decision problem version of top-k planning is PSPACE-complete, as is the decision problem version of ordinary classical planning. This does not hold for polynomially bounded plans for which the decision problem turns out to be PP-hard, while the ordinary case is NP-hard. We present a novel approach to top-k planning, called SYM-K, which is based on symbolic search, and prove that SYM-K is sound and complete. Our empirical analysis shows that SYM-K exceeds the current state of the art for both small and large k.

• Daniel Reifsteck, Thorsten Engesser, Robert Mattmüller and Bernhard Nebel.
  Epistemic Multi-agent Planning Using Monte-Carlo Tree Search.
  In KI 2019: Advances in Artificial Intelligence - 42nd German Conference on AI, pp. 277-289. 2019.
  (Show abstract) (Hide abstract) (Online; DOI)
  

  Coordination in multi-agent systems with partial and non-uniform observability is a practically challenging problem. We use Monte-Carlo tree search as the basis of an implicitly coordinated epistemic planning algorithm which is capable of using the knowledge distributed among the agents to find solutions in problems even with a large branching factor. We use Dynamic Epistemic Logic to represent the knowledge and the actual situation as a state of the Monte-Carlo tree search, and epistemic planning to formalize the goals and actions of a problem. Further, we describe the required modifications of the Monte-Carlo tree search when searching over epistemic states, and make use of the cooperative card game Hanabi to test our planner on larger problems. We find that the approach scales to games with up to eight cards while maintaining high playing strength.

• Sumitra Corraya, Florian Geißer, David Speck and Robert Mattmüller.
  An Empirical Study of the Usefulness of State-Dependent Action Costs in Planning.
  In Proceedings of the 42nd German Conference on Artificial Intelligence (KI 2019), pp. 123-130. 2019.
  (Show abstract) (Hide abstract) (PDF) (PDF; Online)
  

  The vast majority of work in planning to date has focused on state-independent action costs. However, if a planning task features state-dependent costs, using a cost model with state-independent costs means either introducing a modeling error, or potentially sacrificing compactness of the model. In this paper, we investigate the conflicting priorities of modeling accuracy and compactness empirically, with a particular focus on the extent of the negative impact of reduced modeling accuracy on (a) the quality of the resulting plans, and (b) the search guidance provided by heuristics that are fed with inaccurate cost models. Our empirical results show that the plan suboptimality introduced by ignoring state-dependent costs can range, depending on the domain, from inexistent to several orders of magnitude. Furthermore, our results show that the impact on heuristic guidance additionally depends strongly on the heuristic that is used, the specifics of how exactly the costs are represented, and whether one is interested in heuristic accuracy, node expansions, or overall runtime savings.

• Bernhard Nebel, Thomas Bolander, Thorsten Engesser, Robert Mattmüller and .
  Implicitly Coordinated Multi-Agent Path Finding under Destination Uncertainty: Success Guarantees and Computational Complexity (Extended Abstract).
  In Proceedings of International Joint Conference on Artificial Intelligence (IJCAI 2019), pp. 6372-6374. 2019.
  

• David Speck, Florian Geißer, Robert Mattmüller and Álvaro Torralba.
  Symbolic Planning with Axioms.
  In Proceedings of the 29th International Conference on Automated Planning and Scheduling (ICAPS 2019), pp. 464-572. 2019.
  (Show abstract) (Hide abstract) (PDF)
  

  Axioms are an extension for classical planning models that allow for modeling complex preconditions and goals exponentially more compactly. Although axioms were introduced in planning more than a decade ago, modern planning techniques rarely support axioms, especially in cost-optimal planning. Symbolic search is a popular and competitive optimal planning technique based on the manipulation of sets of states. In this work, we extend symbolic search algorithms to support axioms natively. We analyze different ways of encoding derived variables and axiom rules to evaluate them in a symbolic representation. We prove that all encodings are sound and complete, and empirically show that the presented approach outperforms the previous state of the art in cost-optimal classical planning with axioms.

• Thomas Bolander, Thorsten Engesser, Andreas Herzig, Robert Mattmüller and Bernhard Nebel.
  The Dynamic Logic of Policies and Contingent Planning.
  In Logics in Artificial Intelligence - 16th European Conference (JELIA-2019), pp. 659-674. 2019.
  (Show abstract) (Hide abstract) (Online; DOI)
  

  In classical deterministic planning, solutions to planning tasks are simply sequences of actions, but that is not sufficient for contingent plans in non-deterministic environments. Contingent plans are often expressed through policies that map states to actions. An alternative is to specify contingent plans as programs, e.g. in the syntax of Propositional Dynamic Logic (PDL). PDL is a logic for reasoning about programs with sequential composition, test and non-deterministic choice. However, as we show in the paper, none of the existing PDL modalities directly captures the notion of a solution to a planning task under non-determinism. We add a new modality to star-free PDL correctly capturing this notion. We prove the appropriateness of the new modality by showing how to translate back and forth between policies and PDL programs under the new modality. More precisely, we show how a policy solution to a planning task gives rise to a program solution expressed via the new modality, and vice versa. We also provide an axiomatisation of our PDL extension through reduction axioms into standard star-free PDL.

• Bernhard Nebel, Thomas Bolander, Thorsten Engesser and Robert Mattmüller.
  Implicitly Coordinated Multi-Agent Path Finding under Destination Uncertainty: Success Guarantees and Computational Complexity.
  Journal of Artificial Intelligence Research 64, pp. 497-527. 2019.
  (Show abstract) (Hide abstract) (PDF)
  

  In multi-agent path finding (MAPF), it is usually assumed that planning is performed centrally and that the destinations of the agents are common knowledge. We will drop both assumptions and analyze under which conditions it can be guaranteed that the agents reach their respective destinations using implicitly coordinated plans without communication. Furthermore, we will analyze what the computational costs associated with such a coordination regime are. As it turns out, guarantees can be given assuming that the agents are of a certain type. However, the implied computational costs are quite severe. In the distributed setting, we either have to solve a sequence of NP-complete problems or have to tolerate exponentially longer executions. In the setting with destination uncertainty, bounded plan existence becomes PSPACE-complete. This clearly demonstrates the value of communicating about plans before execution starts.

• Felix Lindner, Robert Mattmüller and Bernhard Nebel.
  Moral Permissibility of Action Plans.
  In Proceedings of the Thirty-Third AAAI Conference on Artificial Intelligence (AAAI-19). 2019.
  (Show abstract) (Hide abstract) (PDF)
  

  Research in classical planning so far was mainly concerned with generating a satisficing or an optimal plan. However, if such systems are used to make decisions that are relevant to humans, one should also consider the ethical consequences generated plans can have. We address this challenge by analyzing in how far it is possible to generalize existing approaches of machine ethics to automatic planning systems. Traditionally, ethical principles are formulated in an action-based manner, allowing to judge the execution of one action. We show how such a judgment can be generalized to plans. Further, we study the computational complexity of making ethical judgment about plans.

• Thomas Bolander, Thorsten Engesser, Robert Mattmüller and Bernhard Nebel.
  Better Eager Than Lazy? How Agent Types Impact the Successfulness of Implicit Coordination.
  In Proceedings of the Sixteenth Conference on Principles of Knowledge Representation and Reasoning (KR18), pp. 445-453. 2018.
  (Show abstract) (Hide abstract) (PDF)
  

  Epistemic planning can be used for decision making in multi-agent situations with distributed knowledge and capabilities. In recent work, we proposed a new notion of strong policies with implicit coordination. With this it is possible to solve planning tasks with joint goals from a single-agent perspective without the agents having to negotiate about and commit to a joint policy at plan time. We study how and under which circumstances the decentralized application of those policies leads to the desired outcome.

• Benedict Wright, Robert Mattmüller and Bernhard Nebel.
  Compiling Away Soft Trajectory Constraints in Planning.
  In Proceedings of the Sixteenth COnference on Principles of Knowledge Representation and Reasoning (KR18), pp. 474-482. 2018.
  (Show abstract) (Hide abstract) (PDF)
  

  Soft goals in planning are optional objectives that should be achieved in the terminal state. However, failing to achieve them does not result in the plan becoming invalid. State trajectory constraints are hard requirements towards the state trajectory of the plan. Soft trajectory constraints are a combination of both: soft preferences on how the hard goals are reached, i. e., optional requirements towards the state trajectory of the plan. Such a soft trajectory constraint may require that some fact should be always true, or should be true at some point during the plan. The quality of a plan is then measured by a metric which adds the sum of all action costs and a penalty for each failed soft trajectory constraint. Keyder and Geffner showed that soft goals can be compiled away. We generalize this approach and illustrate a method of compiling soft trajectory constraints into conditional effects and state-dependent action costs using LTL f and deterministic finite automata. We provide two compilation schemes, with and without reward shaping, by rewarding and penalizing different states in the plan. With this we are able to handle such soft trajectory constraints without the need of altering the search algorithm or heuristics, using classical planners.

• Thorsten Engesser, Robert Mattmüller, Bernhard Nebel and Michael Thielscher.
  Game Description Language and Dynamic Epistemic Logic Compared.
  In Proceedings of International Joint Conference on Artificial Intelligence (IJCAI 2018), pp. 1795-1802. 2018.
  (Show abstract) (Hide abstract) (PDF)
  

  Several different frameworks have been proposed to model and reason about knowledge in dynamic multi-agent settings, among them the logic-programming-based game description language GDL-III, and dynamic epistemic logic (DEL), based on possible-worlds semantics. GDL-III and DEL have complementary strengths and weaknesses in terms of ease of modeling and simplicity of semantics. In this paper, we formally study the expressiveness of GDL-III vs. DEL. We clarify the commonalities and differences between those languages, demonstrate how to bridge the differences where possible, and identify large fragments of GDL-III and DEL that are equivalent in the sense that they can be used to encode games or planning tasks that admit the same legal action sequences. We prove the latter by providing compilations between those fragments of GDL-III and DEL.

• David Speck, Florian Geißer and Robert Mattmüller.
  Symbolic Planning with Edge-Valued Multi-Valued Decision Diagrams.
  In Proceedings of the 28th International Conference on Automated Planning and Scheduling (ICAPS 2018). 2018.
  (Show abstract) (Hide abstract) (PDF) (technical report with proofs; PDF)
  

  Symbolic representations have attracted significant attention in optimal planning. Binary Decision Diagrams (BDDs) form the basis for symbolic search algorithms. Closely related are Algebraic Decision Diagrams (ADDs), used to represent heuristic functions.

  Also, progress was made in dealing with models that take state-dependent action costs into account. Here, costs are represented as Edge-valued Multi-valued Decision Diagrams (EVMDDs), which can be exponentially more compact than the corresponding ADD representation. However, they were not yet considered for symbolic planning.

  In this work, we study EVMDD-based symbolic search for optimal planning. We define EVMDD-based representations of state sets and transition relations, and show how to compute the necessary operations required for EVMDD-A*. This EVMDD-based version of symbolic A* generalizes its BDD variant, and allows to solve planning tasks with state-dependent action costs.

  We prove theoretically that our approach is sound, complete and optimal. Additionally, we present an empirical analysis comparing EVMDD-A* to BDD-A* and explicit A* search. Our results underscore the usefulness of symbolic approaches and the feasibility of dealing with models that go beyond unit costs.

• Benedict Wright, Robert Mattmüller and Bernhard Nebel.
  Compiling Away Soft Trajectory Constraints in Planning.
  In Proceedings of the Workshop on Knowledge Engineering for Planning and Scheduling (KEPS18), pp. 38-45. 2018.
  (Show abstract) (Hide abstract) (PDF)
  

  Soft goals in planning are optional objectives that should be achieved in the terminal state. However, failing to achieve them does not result in the plan becoming invalid. State trajectory constraints are hard requirements towards the state trajectory of the plan. Soft trajectory constraints are a combination of both: soft preferences on how the hard goals are reached, i. e., optional requirements towards the state trajectory of the plan. Such a soft trajectory constraint may require that some fact should be always true, or should be true at some point during the plan. The quality of a plan is then measured by a metric which adds the sum of all action costs and a penalty for each failed soft trajectory constraint. Keyder and Geffner showed that soft goals can be compiled away. We generalize this approach and illustrate a method of compiling soft trajectory constraints into conditional effects and state-dependent action costs using LTL f and deterministic finite automata. We provide two compilation schemes, with and without reward shaping, by rewarding and penalizing different states in the plan. With this we are able to handle such soft trajectory constraints without the need of altering the search algorithm or heuristics, using classical planners.

• Felix Lindner, Robert Mattmüller and Bernhard Nebel.
  Moral Permissibility of Action Plans.
  In Proceedings of the ICAPS Workshop on EXplainable AI Planning (XAIP). 2018.
  (Show abstract) (Hide abstract) (PDF)
  

  Research in classical planning so far was mainly concerned with generating a satisficing or an optimal plan. However, if such systems are used to make decisions that are relevant to humans, one should also consider the ethical consequences that generated plans can have. We address this challenge by analyzing in how far it is possible to generalize existing approaches of machine ethics to automatic planning systems. Traditionally, ethical principles are formulated in an action-based manner, allowing to judge the execution of one action. We show how such a judgment can be generalized to plans. Further, we study the complexity of making ethical judgment about plans.

• David Speck, Florian Geißer and Robert Mattmüller.
  SYMPLE: Symbolic Planning based on EVMDDs.
  In The 9th International Planning Competition (IPC 2018), pp. 82-85. 2018.
  (Show abstract) (Hide abstract) (PDF)
  

  SYMPLE is a classical planner which performs bidirectional symbolic search. Symbolic search has proven to be a useful approach to optimal classical planning and is usually based on Binary Decision Diagrams (BDDs). Our approach is based on an alternative data structure called Edge-valued Multi-valued Decision Diagrams (EVMDDs), which have some structural advantages over BDDs.

• Robert Mattmüller, Florian Geißer, Benedict Wright and Bernhard Nebel.
  On the Relationship Between State-Dependent Action Costs and Conditional Effects in Planning.
  In Proceedings of the Thirty-Second AAAI Conference on Artificial Intelligence (AAAI 2018). 2018.
  (Show abstract) (Hide abstract) (PDF)
  

  When planning for tasks that feature both state-dependent action costs and conditional effects using relaxation heuristics, the following problem appears: handling costs and effects separately leads to worse-than-necessary heuristic values, since we may get the more useful effect at the lower cost by choosing different values of a relaxed variable when determining relaxed costs and relaxed active effects.

  In this paper, we show how this issue can be avoided by representing state-dependent costs and conditional effects uniformly, both as edge-valued multi-valued decision diagrams (EVMDDs) over different sets of edge values, and then working with their product diagram. We develop a theory of EVMDDs that is general enough to encompass state-dependent action costs, conditional effects, and even their combination.

  We define relaxed effect semantics in the presence of state-dependent action costs and conditional effects, and describe how this semantics can be efficiently computed using product EVMDDs. This will form the foundation for informative relaxation heuristics in the setting with state-dependent costs and conditional effects combined.

• Benedict Wright and Robert Mattmüller.
  Automated Data Management Workflow Generation with Ontologies and Planning.
  In Proceedings of the 30th Workshop on Planen/Scheduling und Konfigurieren/Entwerfen (PUK 2016). 2016.
  (PDF)
  

• Robert Mattmüller, Florian Geißer, Benedict Wright and Bernhard Nebel.
  On the Relationship Between State-Dependent Action Costs and Conditional Effects in Planning.
  In Proceedings of the 9th Workshop on Heuristics and Search for Domain-Independent Planning (HSDIP 2017). 2017.
  Superseded by the AAAI 2018 paper by the same name.
  (Show abstract) (Hide abstract) (PDF)
  

  When planning for tasks that feature both state-dependent action costs and conditional effects using relaxation heuristics, the following problem appears: handling costs and effects separately leads to worse-than-necessary heuristic values, since we may get the more useful effect at the lower cost by choosing different values of a relaxed variable when determining relaxed costs and relaxed active effects.

  In this paper, we show how this issue can be avoided by representing state-dependent costs and conditional effects uniformly, both as edge-valued multi-valued decision diagrams (EVMDDs) over different sets of edge values, and then working with their product diagram. We develop a theory of EVMDDs that is general enough to encompass state-dependent action costs, conditional effects, and even their combination.

  We define relaxed effect semantics in the presence of state-dependent action costs and conditional effects, and describe how this semantics can be efficiently computed using product EVMDDs. This will form the foundation for informative relaxation heuristics in the setting with state-dependent costs and conditional effects combined.

• Thorsten Engesser, Thomas Bolander, Robert Mattmüller and Bernhard Nebel.
  Cooperative Epistemic Multi-Agent Planning for Implicit Coordination.
  In Ghosh, Sujata and Ramanujam and R. (eds.), Proceedings of the Ninth Workshop on Methods for Modalities (M4M 2017), pp. 75-90. 2017.
  (Show abstract) (Hide abstract) (Online; PDF) (BIB)
  

  Epistemic planning can be used for decision making in multi-agent situations with distributed knowledge and capabilities. Recently, Dynamic Epistemic Logic (DEL) has been shown to provide a very natural and expressive framework for epistemic planning. We extend the DEL-based epistemic planning framework to include perspective shifts, allowing us to define new notions of sequential and conditional planning with implicit coordination. With these, it is possible to solve planning tasks with joint goals in a decentralized manner without the agents having to negotiate about and commit to a joint policy at plan time. First we define the central planning notions and sketch the implementation of a planning system built on those notions. Afterwards we provide some case studies in order to evaluate the planner empirically and to show that the concept is useful for multi-agent systems in practice.

• Thomas Keller, Florian Pommerening, Jendrik Seipp, Florian Geißer and Robert Mattmüller.
  State-dependent Cost Partitionings for Cartesian Abstractions in Classical Planning (Extended Abstract).
  In Proceedings of the 39th German Conference on Artificial Intelligence (KI 2016). 2016.
  (Show abstract) (Hide abstract) (PDF)
  

  Abstraction heuristics are a popular method to guide optimal search algorithms in classical planning. Cost partitionings allow to sum heuristic estimates admissibly by partitioning action costs among the abstractions. We introduce state-dependent cost partitionings which take context information of actions into account, and show that an optimal state-dependent cost partitioning dominates its state-independent counterpart. We demonstrate the potential of state-dependent cost partitionings with a state-dependent variant of the recently proposed saturated cost partitioning, and show that it can sometimes improve not only over its state-independent counterpart, but even over the optimal state-independent cost partitioning.

• Thomas Keller, Florian Pommerening, Jendrik Seipp, Florian Geißer and Robert Mattmüller.
  State-dependent Cost Partitionings for Cartesian Abstractions in Classical Planning.
  In Proceedings of the 25th International Joint Conference on Artificial Intelligence (IJCAI 2016). 2016.
  (Show abstract) (Hide abstract) (PDF)
  

  Abstraction heuristics are a popular method to guide optimal search algorithms in classical planning. Cost partitionings allow to sum heuristic estimates admissibly by distributing action costs among the heuristics. We introduce state-dependent cost partitionings which take context information of actions into account, and show that an optimal state-dependent cost partitioning dominates its state-independent counterpart. We demonstrate the potential of our idea with a state-dependent variant of the recently proposed saturated cost partitioning, and show that it has the potential to improve not only over its state-independent counterpart, but even over the optimal state-independent cost partitioning. Our empirical results give evidence that ignoring the context of actions in the computation of a cost partitioning leads to a significant loss of information.

• Florian Geißer, Thomas Keller and Robert Mattmüller.
  Abstractions for Planning with State-Dependent Action Costs.
  In Proceedings of the 26th International Conference on Automated Planning and Scheduling (ICAPS 2016). 2016.
  (Show abstract) (Hide abstract) (PDF)
  

  Extending the classical planning formalism with state-dependent action costs (SDAC) allows an up to exponentially more compact task encoding. Recent work proposed to use edge-valued multi-valued decision diagrams (EVMDDs) to represent cost functions, which allows to automatically detect and exhibit structure in cost functions and to make heuristic estimators accurately reflect SDAC. However, so far only the inadmissible additive heuristic has been considered in this context. In this paper, we define informative admissible abstraction heuristics which enable optimal planning with SDAC. We discuss how abstract cost values can be extracted from EVMDDs that represent concrete cost functions without adjusting them to the selected abstraction. Our theoretical analysis shows that this is efficiently possible for abstractions that are Cartesian or coarser. We adapt the counterexample-guided abstraction refinement approach to derive such abstractions. An empirical evaluation of the resulting heuristic shows that highly accurate values can be computed quickly.

• Thomas Bolander, Thorsten Engesser, Robert Mattmüller and Bernhard Nebel.
  Better Eager Than Lazy? How Agent Types Impact the Successfulness of Implicit Coordination.
  In Proceedings of the ICAPS-2016 Workshop on Distributed and Multi-Agent Planning (DMAP 2016). 2016.
  Superseded by the KR18 paper by the same authors.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  Epistemic planning can be used for decision making in multi-agent situations with distributed knowledge and capabilities. Recent work proposed a new notion of strong policies with implicit coordination. With this it is possible to solve planning tasks with joint goals from a single-agent perspective without the agents having to negotiate about and commit to a joint policy at plan time. We study how and under which circumstances the decentralized application of those policies leads to the desired outcome.

• Johannes Aldinger, Robert Mattmüller and Moritz Göbelbecker.
  Complexity Issues of Interval Relaxed Numeric Planning.
  In KI 2015: Advances in Artificial Intelligence (KI 2015). 2015.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  Automated planning is computationally hard even in its most basic form as STRIPS planning. We are interested in numeric planning with instantaneous actions, a problem that is not decidable in general. Relaxation is an approach to simplifying complex problems in order to obtain guidance in the original problem. We present a relaxation approach with intervals for numeric planning and show that plan existence can be decided in polynomial time for tasks where dependencies between numeric effects are acyclic.

• David Speck, Manuela Ortlieb and Robert Mattmüller.
  Necessary Observations in Nondeterministic Planning.
  In Proceedings of the 38th German Conference on Artificial Intelligence (KI 2015). 2015.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  An agent that interacts with a nondeterministic environment can often only partially observe the surroundings. This necessitates observations via sensors rendering more information about the current world state. Sensors can be expensive in many regards therefore it can be essential to minimize the amount of sensors an agent requires to solve given tasks. A limitation for sensor minimization is given by essential sensors which are always required to solve particular problems. In this paper we present an efficient algorithm which determines a set of necessary observation variables. More specifically, we develop a bottom-up algorithm which computes a set of variables which are always necessary to observe, in order to always reach a goal state. Our experimental results show that the knowledge about necessary observation variables can be used to minimize the number of sensors of an agent.

• Florian Geißer, Thomas Keller and Robert Mattmüller.
  Delete Relaxations for Planning with State-Dependent Action Costs.
  In Proceedings of the 24th International Joint Conference on Artificial Intelligence (IJCAI 2015). 2015.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  Most work in planning focuses on tasks with state-independent or even uniform action costs. However, supporting state-dependent action costs admits a more compact representation of many tasks. We investigate how to solve such tasks using heuristic search, with a focus on delete-relaxation heuristics. We first define a generalization of the additive heuristic to such tasks and then discuss different ways of computing it via compilations to tasks with state-independent action costs and more directly by modifying the relaxed planning graph. We evaluate these approaches theoretically and present an implementation of the generalized additive heuristic for planning with state-dependent action costs. To our knowledge, this gives rise to the first approach able to handle even the hardest instances of the combinatorial Academic Advising domain from the International Probabilistic Planning Competition (IPPC) 2014.

• Florian Geißer, Thomas Keller and Robert Mattmüller.
  Delete Relaxations for Planning with State-Dependent Action Costs.
  In Proceedings of the 8th Annual Symposium on Combinatorial Search (SoCS 2015). 2015.
  Extended abstract of the IJCAI 2015 paper by the same name.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  Supporting state-dependent action costs in planning admits a more compact representation of many tasks. We generalize the additive heuristic and compute it by embedding decision-diagram representations of action cost functions into the RPG. We give a theoretical evaluation and present an implementation of the generalized additive heuristic. This allows us to handle even the hardest instances of the combinatorial Academic Advising domain from the IPPC 2014.

• Johannes Aldinger, Robert Mattmüller and Moritz Göbelbecker.
  Complexity Issues of Interval Relaxed Numeric Planning.
  In Proceedings of the ICAPS-2015 Workshop on Heuristic and Search for Domain-Independent Planning (HSDIP 2015). 2015.
  Superseded by the KI 2015 paper of the same name..
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  Automated planning is a hard problem even in its most basic form as STRIPS planning. We are interested in numeric planning tasks with instantaneous actions, a problem which is not even decidable in general. Relaxation is an approach to simplifying complex problems in order to obtain guidance in the original problem. In this paper we present a relaxation approach with intervals for numeric planning and discuss the arising complexity issues.

• Thorsten Engesser, Thomas Bolander, Robert Mattmüller and Bernhard Nebel.
  Cooperative Epistemic Multi-Agent Planning With Implicit Coordination.
  In Proceedings of the ICAPS-2015 Workshop on Distributed and Multi-Agent Planning (DMAP 2015). 2015.
  Superseded by the M4M 2017 paper by the same authors.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  Epistemic Planning has been used to achieve ontic and epistemic control in multi-agent situations. We extend the formalism to include perspective shifts, allowing us to define a class of cooperative problems in which both action planning and execution is done in a purely distributed fashion, meaning coordination is only allowed implicitly by means of the available epistemic actions. While this approach can be fruitfully applied to model reasoning in some simple social situations, we also provide some benchmark applications to show that the concept is useful for multi-agent systems in practice.

• Jonas Thiem, Robert Mattmüller and Manuela Ortlieb.
  Counterexample-Guided Abstraction Refinement for POND Planning.
  In Proceedings of the ICAPS-2015 Workshop on Model Checking and Automated Planning (MOCHAP 2015). 2015.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  Counterexample-guided abstraction refinement (CEGAR) allows to gradually refine a problem definition until the required detail for a solution is present. We propose the use of CEGAR to demonstrate unsolvability of a planning problem while avoiding a search through the whole state space. This allows a sensor minimization algorithm for partially observable non-deterministic (POND) planning problems to determine the definitive unsolvability notably faster, which is a necessary part of current approaches of computing the minimal sensor variable set. We determine from our empirical results that while the presented algorithm causes some slowdown for solvable problems, the unsolvability is determined at a much greater speed with this novel approach.

• Dominik Winterer, Robert Mattmüller and Martin Wehrle.
  Stubborn Sets for Fully Observable Nondeterministic Planning.
  In Proceedings of the ICAPS-2015 Workshop on Model Checking and Automated Planning (MOCHAP 2015). 2015.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  The stubborn set method is a state-space reduction technique, originally introduced in model checking and then transfered to classical planning. It was shown that stubborn sets significantly improve the performance of optimal deterministic planners by considering only a subset of applicable operators in a state. Fully observable nondeterministic planning (FOND) extends the formalism of classical planning by nondeterministic operators. We show that stubborn sets are also beneficial for FOND problems. We introduce nondeterministic stubborn sets, stubborn sets which preserve strong cyclic plans. We follow two approaches: Fast Incremental Planning with stubborn sets from classical planning and LAO* search with nondeterministic stubborn sets. Our experiments show that both approaches increase coverage and decrease node generations when compared to their respective baselines.

• Robert Mattmüller, Manuela Ortlieb and Erik Wacker.
  Minimizing Necessary Observations for Nondeterministic Planning.
  In Proceedings of the 37th German Conference on Artificial Intelligence (KI 2014). 2014.
  (Show abstract) (Hide abstract) (PDF)
  

  Autonomous agents interact with their environments via sensors and actuators. Motivated by the observation that sensors can be expensive, in this paper we are concerned with the problem of minimizing the amount of sensors an agent needs in order to successfully plan and act in a partially observable nondeterministic environment. More specifically, we present a simple greedy top-down algorithm in the space of observation variables that returns an inclusion minimal set of state variables sufficient to observe in order to find a plan. We enhance the algorithm by reusing plans from earlier iterations and by the use of functional dependencies between variables that allows the values of some variables to be inferred from those of other variables. Our experimental evaluation on a number of benchmark problems shows promising results regarding runtime, numbers of sensors and plan quality.

• Andreas Hertle, Christian Dornhege, Thomas Keller, Robert Mattmüller, Manuela Ortlieb and Bernhard Nebel.
  An Experimental Comparison of Classical, FOND and Probabilistic Planning.
  In Proceedings of the 37th German Conference on Artificial Intelligence (KI 2014), pp. 297-308. Springer 2014.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  Domain-independent planning in general is broadly applicable to a wide range of tasks. Many formalisms exist that allow to describe different aspects of realistic problems. Which one is best is not clear as usually more expressiveness comes with a cost in planning time. Under the assumption that hard guarantees are not required, users are faced with a decision between multiple approaches. In this work we study the effect of nondeterministic action outcomes. As a generic model we use a probabilistic description in the form of Markov Decision Processes (MDPs). We define abstracting translations into a classical planning formalism and fully observable nondeterministic (FOND) planning. Our goal is to give insight into how state-of-the-art systems perform on different MDP planning domains. We evaluate those MDPs by running state-of-the-art planning systems on the abstracted formalisms.

• Florian Geißer, Thomas Keller and Robert Mattmüller.
  Past, Present, and Future: An Optimal Online Algorithm for Single-Player GDL-II Games.
  In Proceedings of the 21st European Conference on Artificial Intelligence (ECAI 2014), pp. 357-362. 2014.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  In General Game Playing, a player receives the rules of an unknown game and attempts to maximize his expected reward. Since 2011, the GDL-II rule language extension allows the formulation of nondeterministic and partially observable games. In this paper, we present an algorithm for such games, with a focus on the single-player case. Conceptually, at each stage, the proposed Norns algorithm distinguishes between the past, present and future steps of the game. More specifically, a belief state tree is used to simulate a potential past that leads to a present that is consistent with received observations. Unlike other related methods, our method is asymptotically optimal. Moreover, augmenting the belief state tree with iteratively improved probabilities speeds up the process over time significantly. As this allows a true picture of the present, we additionally present an optimal version of the well-known UCT algorithm for partially observable single-player games. Instead of performing hindsight optimization on a simplified, fully observable tree, the true future is simulated on an action-observation tree that takes partial observability into account. The expected reward estimates of applicable actions converge towards the true expected rewards even for moves that are only used to gather information. We prove that our algorithm is asymptotically optimal for single-player games and POMDPs and support our claim with an empirical evaluation.

• Dr. Yusra Alkhazraji, Michael Katz, Robert Mattmüller, Florian Pommerening, Alexander Shleyfman and Martin Wehrle.
  Metis: Arming Fast Downward with Pruning and Incremental Computation (planner abstract).
  In the 8th International Planning Competition (IPC 2014) (deterministic track). 2014.
  (Show abstract) (Hide abstract) (PDF)
  

  Metis is a sequential optimal planner that implements three components on top of the Fast Downward planning system (Helmert 2006). The planner performs an A ∗ search using the following three major components: • an admissible incremental LM-cut heuristic (Pommeren- ing and Helmert 2013), • a symmetry based pruning technique (Domshlak, Katz, and Shleyfman 2012), and • a partial order reduction based pruning technique based on strong stubborn sets (Wehrle and Helmert 2012). Each of those techniques was extended to support conditional effects. In addition, Metis features a flexible invocation of partial order reduction based pruning. In what follows, we describe each of these components in detail.

• Manuela Ortlieb and Robert Mattmüller.
  Pattern-Database Heuristics for Partially Observable Nondeterministic Planning.
  In Proceedings of the 36th German Conference on Artificial Intelligence (KI 2013). 2013.
  (Show abstract) (Hide abstract) (PDF) (slides; PDF)
  

  Heuristic search is the dominant approach to classical planning. However, many realistic problems violate classical assumptions such as determinism of action outcomes or full observability. In this paper, we investigate how - and how successfully - a particular classical technique, namely informed search using an abstraction heuristic, can be transferred to nondeterministic planning under partial observability. Specifically, we explore pattern-database heuristics with automatically generated patterns in the context of informed progression search for strong cyclic planning under partial observability. To that end, we discuss projections and how belief states can be heuristically assessed either directly or by going back to the contained world states, and empirically evaluate the resulting heuristics internally and compared to a delete-relaxation and a blind approach. From our experiments we can conclude that in terms of guidance, it is preferable to represent both nondeterminism and partial observability in the abstraction (instead of relaxing them), and that the resulting abstraction heuristics significantly outperform both blind search and a delete-relaxation approach where nondeterminism and partial observability are also relaxed.

• Martin Wehrle, Malte Helmert, Dr. Yusra Alkhazraji and Robert Mattmüller.
  The Relative Pruning Power of Strong Stubborn Sets and Expansion Core.
  In Proceedings of the 23rd International Conference on Automated Planning and Scheduling (ICAPS13). 2013.
  (Show abstract) (Hide abstract) (PDF)
  

  In the last years, pruning techniques based on partial order reduction have found increasing attention in the planning community. One recent result is that the expansion core method is a special case of the strong stubborn sets method proposed in model checking. However, it is still an open question if there exist efficiently computable strong stubborn sets with strictly higher pruning power than expansion core. In this paper, we prove that the pruning power of strong stubborn sets is strictly higher than the pruning power of expansion core even for a straight-forward instantiation of strong stubborn sets. This instantiation is as efficiently computable as expansion core. Hence, our theoretical results suggest that strong stubborn sets should be preferred to expansion core. Our empirical evaluation on all optimal benchmarks from the international planning competitions up to 2011 supports the theoretical results.

• Dr. Yusra Alkhazraji, Martin Wehrle, Robert Mattmüller and Malte Helmert.
  A Stubborn Set Algorithm for Optimal Planning.
  In Proceedings of the 20th European Conference on Artificial Intelligence (ECAI 2012). 2012.
  (Show abstract) (Hide abstract) (PDF)
  

  We adapt a partial order reduction technique based on stubborn sets, originally proposed for detecting dead ends in Petri Nets, to the setting of optimal planning. We demonstrate that stubborn sets can provide significant state space reductions on standard planning benchmarks, outperforming the expansion core method.

• Hans-Jörg Peter, Rüdiger Ehlers and Robert Mattmüller.
  Synthia: Verification and Synthesis for Timed Automata.
  In Proceedings of the 23rd International Conference on Computer Aided Verification (CAV 2011), pp. 649-655. Springer-Verlag 2011.
  (Show abstract) (Hide abstract) (PDF)
  

  We present Synthia, a new tool for the verification and synthesis of open real-time systems modeled as timed automata. The key novelty of Synthia is the underlying abstraction refinement approach that combines the efficient symbolic treatment of timing information by difference bound matrices (DBMs) with the usage of binary decision diagrams (BDDs) for the discrete parts of the system description. Our experiments show that the analysis of both closed and open systems greatly benefits from identifying large relevant and irrelevant system parts on coarse abstractions early in the solution process. Synthia is licensed under the GNU GPL and available from our website.

• Rüdiger Ehlers, Robert Mattmüller and Hans-Jörg Peter.
  Combining Symbolic Representations for Solving Timed Games.
  In Proceedings of the 8th International Conference on Formal Modelling and Analysis of Timed Systems (FORMATS 2010), pp. 107-121. Springer-Verlag 2010.
  (Show abstract) (Hide abstract) (PDF)
  

  We present a general approach to combine symbolic state space representations for the discrete and continuous parts in the synthesis of winning strategies for timed reachability games. The combination is based on abstraction refinement where discrete symbolic techniques are used to produce a sequence of abstract timed game automata. After each refinement step, the resulting abstraction is used for computing an under- and an over-approximation of the timed winning states. The key idea is to identify large relevant and irrelevant parts of the precise weakest winning strategy already on coarse, and therefore simple, abstractions. If neither the existence nor nonexistence of a winning strategy can be established in the approximations, we use them to guide the refinement process. Based on a prototype that combines binary decision diagrams and difference bound matrices, we experimentally evaluate the technique on standard benchmarks from timed controller synthesis. The results clearly demonstrate the potential of the new approach concerning running time and memory consumption compared to the classical on-the-fly algorithm implemented in UPPAAL-Tiga.

• J. Benton, Kartik Talamadupula, Patrick Eyerich, Robert Mattmüller and Subbarao Kambhampati.
  G-value Plateaus: A Challenge for Planning.
  In Ronen Brafman, Héctor Geffner, Jörg Hoffmann and Henry Kautz (eds.), Proceedings of the 20th International Conference on Automated Planning and Scheduling (ICAPS 2010), pp. 259-262. AAAI Press 2010.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  Recent years have seen the development of several scalable planners, many of which follow the string of successes found in using heuristic, best-first search methods. While this provides positive reinforcement for continuing work along these lines, fundamental problems arise when handling objectives whose value does not change with each search operation. An extreme case of this occurs when handling the objective of generating a temporal plan with short makespan. Typically used heuristic search methods assume strictly positive edge costs for their guarantees on completeness and optimality to hold, while the usual "fattening" and "advance time" steps of heuristic search algorithms for temporal planning have the potential for zero-cost edges, resulting in "g-value plateaus". In this paper we point out some underlying difficulties with using modern heuristic search methods for optimizing makespan and discuss how the presence of these problems contributes to the poor performance of makespan-optimizing heuristic search planners. To further illustrate this, we show empirical results on recent benchmarks using a planner made with makespan optimization in mind.

• Robert Mattmüller, Manuela Ortlieb, Malte Helmert and Pascal Bercher.
  Pattern Database Heuristics for Fully Observable Nondeterministic Planning.
  In Ronen Brafman, Héctor Geffner, Jörg Hoffmann and Henry Kautz (eds.), Proceedings of the 20th International Conference on Automated Planning and Scheduling (ICAPS 2010), pp. 105-112. AAAI Press 2010.
  (Show abstract) (Hide abstract) (PDF) (slides; PDF) (BIB)
  

  When planning in an uncertain environment, one is often interested in finding a contingent plan that prescribes appropriate actions for all possible states that may be encountered during the execution of the plan. We consider the problem of finding strong and strong cyclic plans for fully observable nondeterministic (FOND) planning problems. The algorithm we choose is LAO*, an informed explicit state search algorithm. We investigate the use of pattern database (PDB) heuristics to guide LAO* towards goal states. To obtain a fully domain-independent planning system, we use an automatic pattern selection procedure that performs local search in the space of pattern collections. The evaluation of our system on the FOND benchmarks of the Uncertainty Part of the International Planning Competition 2008 shows that our approach is competitive with symbolic regression search in terms of problem coverage and speed.

• Hans-Jörg Peter and Robert Mattmüller.
  Component-based Abstraction Refinement for Timed Controller Synthesis.
  In Theodore P. Baker (ed.), Proceedings of the 30th IEEE Real-Time Systems Symposium (RTSS 2009), pp. 364-374. IEEE Computer Society 2009.
  (Show abstract) (Hide abstract) (PDF) (BIB)
  

  We present a novel technique for synthesizing controllers for distributed real-time environments with safety requirements. Our approach is an abstraction refinement extension to the on-the-fly algorithm by Cassez et al. from 2005. Based on partial compositions of some environment components, each refinement cycle constructs a sound abstraction that can be used to obtain under- and over-approximations of all valid controller implementations. This enables (1) early termination if an implementation does not exist in the over-approximation, or, if one does exist in the under-approximation, and (2) pruning of irrelevant moves in subsequent refinement cycles. In our refinement loop, the precision of the abstractions incrementally increases and converges to all specification-critical components.

  We implemented our approach in a prototype synthesis tool and evaluated it on an industrial benchmark. In comparison with the timed game solver UPPAAL-Tiga, our technique outperforms the nonincremental approach by an order of magnitude.

• Patrick Eyerich, Robert Mattmüller and Gabriele Röger.
  Using the Context-enhanced Additive Heuristic for Temporal and Numeric Planning.
  In Proceedings of the 19th International Conference on Automated Planning and Scheduling (ICAPS 2009), pp. 130-137. AAAI Press 2009.
  (Show abstract) (Hide abstract) (PDF) (slides; PDF) (BIB)
  

  Planning systems for real-world applications need the ability to handle concurrency and numeric fluents. Nevertheless, the predominant approach to cope with concurrency followed by the most successful participants in the latest International Planning Competitions (IPC) is still to find a sequential plan that is rescheduled in a post-processing step. We present Temporal Fast Downward (TFD), a planning system for temporal problems that is capable of finding low-makespan plans by performing a heuristic search in a temporal search space. We show how the context-enhanced additive heuristic can be successfully used for temporal planning and how it can be extended to numeric fluents. TFD often produces plans of high quality and, evaluated according to the rating scheme of the last IPC, outperforms all state-of-the-art temporal planning systems.

• Pascal Bercher and Robert Mattmüller.
  Solving Non-deterministic Planning Problems with Pattern Database Heuristics.
  In B. Mertsching, M. Hund and Z. Aziz (eds.), Proceedings of the 32nd Annual Conference on Artificial Intelligence (KI 2009), pp. 57-64. Springer-Verlag 2009.
  (Show abstract) (Hide abstract) (PDF) (slides; PDF) (BIB)
  

  Non-determinism arises naturally in many real-world applications of action planning. Strong plans for this type of problems can be found using AO* search guided by an appropriate heuristic function. Most domain-independent heuristics considered in this context so far are based on the idea of ignoring delete lists and do not properly take the non-determinism into account. Therefore, we investigate the applicability of pattern database (PDB) heuristics to non-deterministic planning. PDB heuristics have emerged as rather informative in a deterministic context. Our empirical results suggest that PDB heuristics can also perform reasonably well in non-deterministic planning. Additionally, we present a generalization of the pattern additivity criterion known from classical planning to the non-deterministic setting.

• Pascal Bercher and Robert Mattmüller.
  A Planning Graph Heuristic for Forward-Chaining Adversarial Planning.
  In Proceedings of the 18th European Conference on Artificial Intelligence (ECAI 2008), pp. 921-922. IOS Press 2008.
  (Show abstract) (Hide abstract) (PDF) (slides; PDF) (poster; PDF) (BIB)
  

  In contrast to classical planning, in adversarial planning, the planning agent has to face an adversary trying to prevent him from reaching his goals. In this paper, we investigate a forward-chaining approach to adversarial planning based on the AO* algorithm. The exploration of the underlying AND/OR graph is guided by a heuristic evaluation function, inspired by the relaxed planning graph heuristic used in the FF planner. Unlike FF, our heuristic uses an adversarial planning graph with distinct proposition and action layers for the protagonist and antagonist. First results suggest that in certain planning domains, our approach yields results competitive with the state of the art.

• Malte Helmert and Robert Mattmüller.
  Accuracy of Admissible Heuristic Functions in Selected Planning Domains.
  In Proceedings of the 23rd AAAI Conference on Artificial Intelligence (AAAI 2008), pp. 938-943. AAAI Press 2008.
  (Show abstract) (Hide abstract) (PDF) (slides; PDF) (BIB)
  

  The efficiency of optimal planning algorithms based on heuristic search crucially depends on the accuracy of the heuristic function used to guide the search. Often, we are interested in domain-independent heuristics for planning. In order to assess the limitations of domain-independent heuristic planning, it appears interesting to analyse the (in)accuracy of common domain-independent planning heuristics in the IPC benchmark domains. For a selection of these domains, we analytically investigate the accuracy of the h⁺ heuristic, the h^m family of heuristics, and certain (additive) pattern database heuristics, compared to the perfect heuristic h^*. Whereas h⁺ and additive pattern database heuristics usually return cost estimates proportional to the true cost, non-additive h^m and non-additive pattern-database heuristics can yield results underestimating the true cost by arbitrarily large factors.

• Malte Helmert and Robert Mattmüller.
  On the Accuracy of Admissible Heuristic Functions in Selected Planning Domains.
  In Proceedings of the ICAPS-2007 Workshop on Heuristics for Domain-independent Planning: Progress, Ideas, Limitations, Challenges. 2007.
  Superseded by the AAAI 2008 paper by the same name.
  (Show abstract) (Hide abstract) (PDF)
  

  The efficiency of optimal planning algorithms based on heuristic search crucially depends on the accuracy of the heuristic function used to guide the search. Often, we are interested in domain-independent heuristics for planning. In assessing the limitations of domain-independent heuristic planning, it appears interesting to analyse the (in)accuracy of common domain-independent planning heuristics in the IPC benchmark domains. For a selection of these domains, we analytically investigate the accuracy of the h⁺ heuristic, the h^k family of heuristics, and certain (additive) pattern database heuristics, compared to the optimal heuristic h^*. Whereas h⁺ and additive pattern database heuristics usually return cost estimates proportional to the true cost, non-additive h^k and non-additive pattern-database heuristics can yield results underestimating the true cost by arbitrarily large factors.

• Robert Mattmüller and Jussi Rintanen.
  Planning for Temporally Extended Goals as Propositional Satisfiability.
  In Proceedings of the 20th International Joint Conference on Artificial Intelligence (IJCAI 2007), pp. 1966-1971. 2007.
  (Show abstract) (Hide abstract) (PDF) (PS.GZ) (poster; PDF) (BIB)
  

  Planning for temporally extended goals (TEGs) expressed as formulae of Linear-time Temporal Logic (LTL) is a proper generalization of classical planning, not only allowing to specify properties of a goal state but of the whole plan execution. Additionally, LTL formulae can be used to represent domain-specific control knowledge to speed up planning. In this paper we extend SAT-based planning for LTL goals (akin to bounded LTL model-checking in verification) to partially ordered plans, thus significantly increasing planning efficiency compared to purely sequential SAT planning. We consider a very relaxed notion of partial ordering and show how planning for LTL goals (without the next-time operator) can be translated into a SAT problem and solved very efficiently. The results extend the practical applicability of SAT-based planning to a wider class of planning problems. In addition, they could be applied to solving problems in bounded LTL model-checking more efficiently.

• Malte Helmert, Robert Mattmüller and Sven Schewe.
  Selective Approaches for Solving Weak Games.
  In Proceedings of the Fourth International Symposium on Automated Technology for Verification and Analysis (ATVA 2006), pp. 200-214. Springer-Verlag 2006.
  (Show abstract) (Hide abstract) (PDF)
  

  Model-checking alternating-time properties has recently attracted much interest in the verification of distributed protocols. While checking the validity of a specification in alternating-time temporal logic (ATL) against an explicit model is cheap (linear in the size of the formula and the model), the problem becomes EXPTIME-hard when symbolic models are considered. Practical ATL model-checking therefore often consumes too much computation time to be tractable.

  In this paper, we describe a novel approach for ATL model-checking, which constructs an explicit weak model-checking game on-the-fly. This game is then evaluated using heuristic techniques inspired by efficient evaluation algorithms for and/or-trees.

  To show the feasibility of our approach, we compare its performance to the ATL model-checking system MOCHA on some practical examples. Using very limited heuristic guidance, we achieve a significant speedup on these benchmarks.

• Malte Helmert, Robert Mattmüller and Gabriele Röger.
  Approximation Properties of Planning Benchmarks.
  In Proceedings of the 17th European Conference on Artificial Intelligence (ECAI 2006), pp. 585-589. 2006.
  (Show abstract) (Hide abstract) (PDF)
  

  For many classical planning domains, the computational complexity of non-optimal and optimal planning is known. However, little is known about the area in between the two extremes of finding some plan and finding optimal plans. In this contribution, we provide a complete classification of the propositional domains from the first four International Planning Competitions with respect to the approximation classes PO, PTAS, APX, poly-APX, and NPO.