FFRob: An Efficient Heuristic for Task and Motion Planning

Caelan Reed Garrett, Tomás Lozano-Pérez, Leslie Pack Kaelbling

Abstract: Manipulation problemsGarrett, Caelan involvingKaelbling, Leslie manyLozano-Perez, Tomas objects present substantial challenges for motion planning algorithms due to the high dimensionality and multi-modality of the search space. Symbolic task planners can efficiently construct plans involving many entities but cannot incorporate the constraints from geometry and kinematics. In this paper, we show how to extend the heuristic ideas from one of the most successful symbolic planners in recent years, the FastForward (FF) planner, to motion planning, and to compute it efficiently. We use a multi-query roadmap structure that can be conditionalized to model different placements of movable objects. The resulting tightly integrated planner is simple and performs efficiently in a collection of tasks involving manipulationManipulation planning of many objects.