We report progress on an adversarial reasoning system, RAPSODI (Rapid Adversarial Planning with Strategic Opponent-Driven Intelligence). RAPSODI consists of two modules, GameMaster and a fast single agent planner. GameMaster refines and expands plans for two or more adversaries by making calls to a planning service provided by the fast single agent local search planner. RAPSODI employs an iterative plan critic process that results in a contingency plan for each agent, based on our best model of their capabilities, assets, and intents. The process iterates as many times as the user wants as long as conflicts can be found. With each iteration agents get "smarter" in the sense that their plans are expanded to handle more possible conflicts with other agents. The approach is fast, practical and tractable. We describe our problem format, a variant of PDDL, and discuss the advantages of our iterative-refinement plan-critic strategy for decision support.
RAPSODI Adversarial Reasoner
2007-03-01
387452 byte
Conference paper
Electronic Resource
English