Flaw Selection Strategies for Value-Directed Planning

A central issue faced by partial-order, cansal-link (POCL) planning systems is how to select which flaw to resolve when generating the refinements of a partial plan. Domain-independent flaw selection strategies have been discussed extensively in the recent literature (Peot ~ Smith 1993; Joslin & Pollack 1994; Schubert & Gerevini 1995). The PYRRHUS planning system is a decision-theoretic extension to POCL planners that finds optimal plans for a class of goal-directed value functions. Although PYRRHUS uses a branch-and-bound algorithm instead of best-first satisficing search, it is faced with the same flaw selection decision as other POCL planners. This paper explains why popular domain-independent flawselection strategies are ineffective within an optimizing framework, and presents two new strategies that exploit the additional value information available to PYaam~s.