Avoiding Dead Ends in Real-Time Heuristic Search

Many systems, such as mobile robots, need to be controlled in real time. Real-time heuristic search is a popular on-line planning paradigm that supports concurrent planning and execution. However, existing methods do not incorporate a notion of safety and we show that they can perform poorly in domains that contain dead-end states from which a goal cannot be reached. We introduce new real-time heuristic search methods that can guarantee safety if the domain obeys certain properties. We test these new methods on two different simulated domains that contain dead ends, one that obeys the properties and one that does not. We find that empirically the new methods provide good performance. We hope this work encourages further efforts to widen the applicability of real-

[1]  Wheeler Ruml,et al.  Real-Time Search in Dynamic Worlds , 2010, SOCS.

[2]  Maxim Likhachev,et al.  D*lite , 2002, AAAI/IAAI.

[3]  Shlomo Zilberstein,et al.  LAO*: A heuristic search algorithm that finds solutions with loops , 2001, Artif. Intell..

[4]  Nathan R. Sturtevant,et al.  Scrubbing During Learning In Real-time Heuristic Search , 2016, J. Artif. Intell. Res..

[5]  Ashish Kapoor,et al.  Fast Safe Mission Plans for Autonomous Vehicles , 2016 .

[6]  Richard E. Korf,et al.  Real-Time Heuristic Search , 1990, Artif. Intell..

[7]  Christian J. Muise,et al.  Traps, Invariants, and Dead-Ends , 2016, ICAPS.

[8]  Shlomo Zilberstein.  Building Strong Semi-Autonomous Systems , 2015, AAAI.

[9]  Kostas E. Bekris,et al.  Greedy but Safe Replanning under Kinodynamic Constraints , 2007, Proceedings 2007 IEEE International Conference on Robotics and Automation.

[10]  Marco Pavone,et al.  Evaluating Trajectory Collision Probability through Adaptive Importance Sampling for Safe Motion Planning , 2016, Robotics: Science and Systems.

[11]  Nathan R. Sturtevant,et al.  Exponential Deepening A* for Real-Time Agent-Centered Search , 2014, SOCS.

[12]  Mausam,et al.  SixthSense: Fast and Reliable Recognition of Dead Ends in MDPs , 2010, AAAI.

[13]  Jur P. van den Berg,et al.  Generalized reciprocal collision avoidance , 2015, Int. J. Robotics Res..

[14]  Jorge A. Baier,et al.  Avoiding and Escaping Depressions in Real-Time Heuristic Search , 2014, J. Artif. Intell. Res..

[15]  Sven Koenig,et al.  Comparing real-time and incremental heuristic search for real-time situated agents , 2009, Autonomous Agents and Multi-Agent Systems.

[16]  Nathan R. Sturtevant,et al.  Online Detection of Dead States in Real-Time Agent-Centered Search , 2013, SOCS.

[17]  Sebastian Thrun,et al.  ARA*: Anytime A* with Provable Bounds on Sub-Optimality , 2003, NIPS.

[18]  Pieter Abbeel,et al.  Safe Exploration in Markov Decision Processes , 2012, ICML.

[19]  Andrew G. Barto,et al.  Learning to Act Using Real-Time Dynamic Programming , 1995, Artif. Intell..

[20]  Alberto Camacho,et al.  From FOND to Robust Probabilistic Planning: Computing Compact Policies that Bypass Avoidable Deadends , 2016, ICAPS.

[21]  Wheeler Ruml,et al.  Achieving Goals Quickly Using Real-time Search: Experimental Results in Video Games , 2015, J. Artif. Intell. Res..

[22]  Nathan R. Sturtevant,et al.  Learning Where You Are Going and from Whence You Came: h- and g-Cost Learning in Real-Time Heuristic Search , 2011, IJCAI.

[23]  Sven Koenig,et al.  Real-time adaptive A* , 2006, AAMAS '06.

[24]  David J. Musliner,et al.  World Modeling for the Dynamic Construction of Real-Time Control Plans , 1995, Artif. Intell..