An Algorithm for Probabilistic Planning

We de ne the probabilistic planning problem in terms of a probability distribution over initial world states, a boolean combination of propositions representing the goal, a probability threshold, and actions whose e ects depend on the execution-time state of the world and on random chance. Adopting a probabilistic model complicates the de nition of plan success: instead of demanding a plan that provably achieves the goal, we seek plans whose probability of success exceeds the threshold. In this paper, we present buridan, an implemented least-commitment planner that solves problems of this form. We prove that the algorithm is both sound and complete. We then explore buridan's e ciency by contrasting four algorithms for plan evaluation, using a combination of analytic methods and empirical experiments. We also describe the interplay between generating plans and evaluating them, and discuss the role of search control in probabilistic planning. We gratefully acknowledge the comments and suggestions of Tony Barrett, Tom Dean, Denise Draper, Mike Erdmann, Keith Golden, Rex Jacobovits, Oren Etzioni, Neal Lesh, Mike Wellman, Mike Williamson, and the anonymous reviewers. This research was funded in part by National Science Foundation Grants IRI-9206733 and IRI-8957302, O ce of Naval Research Grant 90-J-1904, and the Xerox Corporation.