Planning as Goal Recognition: Deriving Heuristics from Intention Models

Giacomo Rosa, Nir Lipovetzky, Sebastian Sardina, Jean Honorio

Track: Long Paper (9 pages including references)

Previous Publication: No, the submission has not been published or accepted at another conference.

Keywords: Planning, Domain-Independent Planning, Classical Planning, Goal Recognition, Heuristic Search, Agile Planner, Satisficing Planner

Abstract:

Classical Planning seeks to find a sequence of actions, a plan, that maps a starting state into one of the goal states.
If a trajectory appears to be leading to the goal, should we prioritise exploring it?
Seminal work in Goal Recognition (GR) has defined GR in terms of a Classical Planning problem, adopting classical solvers and heuristics to recognise plans. We come full circle, and study the adoption and properties of GR-derived heuristics for seeking solutions to Classical Planning problems. We propose a new class of efficiently-computable heuristics and show that they improve the performance of top-scoring planners. Our work provides foundational knowledge for understanding and deriving probabilistic heuristics for Planning.

Submission Number: 11