Extending PDDL for Hierarchical Planning and Topological Abstraction

Despite major progress in AI planning over the last few years, many interesting domains remain challenging for current planners. Topological abstraction can reduce planning complexity in several domains, decomposing a problem into a two-level hierarchy. This paper presents LAP, a planning model based on topological abstraction. In formalizing LAP as a generic planning framework, the support of a planning language more expressive than PDDL can be very important. We discuss how an extended version of PDDL can be part of our planning framework, by providing support for hierarchical planning and topological abstraction. We demonstrate our ideas in Sokoban and path-finding, two domains where topological abstraction is useful.