Robust planning with incomplete domain models

Most current planners assume complete domain models and focus on generating correct plans. Unfortunately, domain modeling is a laborious and error-prone task, thus real world agents have to plan with incomplete domain models. While domain experts cannot guarantee completeness, often they are able to circumscribe the incompleteness of the model by providing annotations as to which parts of the domain model may be incomplete. In this paper, we study planning problems with incomplete domain models where the annotations specify possible preconditions and effects of actions. We show that the problem of assessing the quality of a plan, or its plan robustness, is #P-complete, establishing its equivalence with the weighted model counting problems. We present two approaches to synthesizing robust plans. While the method based on the compilation to conformant probabilistic planning is much intuitive, its performance appears to be limited to only small problem instances. Our second approach based on stochastic heuristic search works well for much larger problems. It aims to use the robustness measure directly for estimating heuristic distance, which is then used to guide the search. Our planning system, PISA, outperforms a state-of-the-art planner handling incomplete domain models in most of the tested domains, both in terms of plan quality and planning time. Finally, we also present an extension of PISA called CPISA that is able to exploit the available of past successful plan traces to both improve the robustness of the synthesized plans and reduce the domain modeling burden.

[1]  Armando Tacchella,et al.  Theory and Applications of Satisfiability Testing , 2003, Lecture Notes in Computer Science.

[2]  Qiang Yang,et al.  Action-model acquisition for planning via transfer learning , 2014, Artif. Intell..

[3]  Ilkka Niemelä,et al.  Parallel Encodings of Classical Planning as Satisfiability , 2004, JELIA.

[4]  Malcolm J. A. Strens,et al.  A Bayesian Framework for Reinforcement Learning , 2000, ICML.

[5]  Eyal Amir,et al.  Learning Partially Observable Deterministic Action Models , 2005, IJCAI.

[6]  Maria Fox,et al.  Exploration of the Robustness of Plans , 2005, AAAI.

[7]  Subbarao Kambhampati,et al.  Model-Lite Case-Based Planning , 2012, AAAI.

[8]  Manuela M. Veloso,et al.  Fault Tolerant Planning: Toward Probabilistic Uncertainty Models in Symbolic Non-Deterministic Planning , 2004, ICAPS.

[9]  Leslie G. Valiant,et al.  The Complexity of Enumeration and Reliability Problems , 1979, SIAM J. Comput..

[10]  Subbarao Kambhampati,et al.  A Heuristic Approach to Planning with Incomplete STRIPS Action Models , 2014, ICAPS.

[11]  Subbarao Kambhampati,et al.  Synthesizing Robust Plans under Incomplete Domain Models , 2011, NIPS.

[12]  Bernhard Nebel,et al.  The FF Planning System: Fast Plan Generation Through Heuristic Search , 2011, J. Artif. Intell. Res..

[13]  Bart Selman,et al.  Model Counting: A New Strategy for Obtaining Good Bounds , 2006, AAAI.

[14]  Daniel Bryce,et al.  Planning and Acting in Incomplete Domains , 2011, ICAPS.

[15]  J. K. Satia,et al.  Markovian Decision Processes with Uncertain Transition Probabilities , 1973, Oper. Res..

[16]  D. Bryce,et al.  Reachability Heuristics for Planning in Incomplete Domains , 2009 .

[17]  Richard Fikes,et al.  STRIPS: A New Approach to the Application of Theorem Proving to Problem Solving , 1971, IJCAI.

[18]  Chelsea C. White,et al.  Markov Decision Processes with Imprecise Transition Probabilities , 1994, Oper. Res..

[19]  Ronen I. Brafman,et al.  Conformant planning via heuristic forward search: A new approach , 2004, Artif. Intell..

[20]  Subbarao Kambhampati,et al.  Refining Incomplete Planning Domain Models Through Plan Traces , 2013, IJCAI.

[21]  Subbarao Kambhampati,et al.  Assessing and Generating Robust Plans with Partial Domain Models , 2010 .

[22]  Qiang Yang,et al.  Learning action models from plan examples using weighted MAX-SAT , 2007, Artif. Intell..

[23]  Leslie G. Valiant,et al.  The Complexity of Computing the Permanent , 1979, Theor. Comput. Sci..

[24]  T. L. McCluskey,et al.  Planning domain definition using GIPO , 2007, The Knowledge Engineering Review.

[25]  Jörg Hoffmann,et al.  Simulated Penetration Testing: From "Dijkstra" to "Turing Test++" , 2015, ICAPS.

[26]  Andrew Coles,et al.  A New Local-Search Algorithm for Forward-Chaining Planning , 2007, ICAPS.

[27]  T. L. McCluskey,et al.  Action Knowledge Acquisition with Opmaker2 , 2009, ICAART.

[28]  Carmel Domshlak,et al.  Probabilistic Planning via Heuristic Forward Search and Weighted Model Counting , 2007, J. Artif. Intell. Res..

[29]  Henry A. Kautz,et al.  Heuristics for Fast Exact Model Counting , 2005, SAT.

[30]  Blai Bonet,et al.  Planning as heuristic search , 2001, Artif. Intell..

[31]  Robert Givan,et al.  Bounded Parameter Markov Decision Processes , 1997, ECP.

[32]  T. L. McCluskey,et al.  Acquiring planning domain models using LOCM , 2013, The Knowledge Engineering Review.

[33]  Subbarao Kambhampati,et al.  Model-lite Planning for the Web Age Masses: The Challenges of Planning with Incomplete and Evolving Domain Models , 2007, AAAI.

[34]  Carmel Domshlak,et al.  Fast Probabilistic Planning through Weighted Model Counting , 2006, ICAPS.

[35]  Andrew Garland,et al.  Plan evaluation with incomplete action descriptions , 2002, AAAI/IAAI.

[36]  David Andre,et al.  Model based Bayesian Exploration , 1999, UAI.

[37]  Bart Selman,et al.  A New Approach to Model Counting , 2005, SAT.

[38]  Scott Sanner,et al.  Solutions to Factored MDPs with Imprecise Transition Probabilities 1 , 2011 .