Backward-Forward Search for Manipulation Planning

TL;DR

This paper introduces the hybrid backward-forward (HBF) planning algorithm for high-dimensional manipulation tasks, effectively generating long plans in cluttered environments by combining backward constraint identification with forward search.

Contribution

The paper presents a novel hybrid backward-forward (HBF) planning algorithm that improves manipulation planning in complex, high-dimensional spaces by integrating backward constraint detection with forward sampling.

Findings

Probabilistically complete planning algorithm.

Effective in cluttered environments with long manipulation plans.

Handles both prehensile and nonprehensile actions.

Abstract

In this paper we address planning problems in high-dimensional hybrid configuration spaces, with a particular focus on manipulation planning problems involving many objects. We present the hybrid backward-forward (HBF) planning algorithm that uses a backward identification of constraints to direct the sampling of the infinite action space in a forward search from the initial state towards a goal configuration. The resulting planner is probabilistically complete and can effectively construct long manipulation plans requiring both prehensile and nonprehensile actions in cluttered environments.