Learning Long-Horizon Action Dependencies in Sampling-Based Bilevel Planning

Bartłomiej Cieślar, Leslie Pack Kaelbling, Tomás Lozano-Pérez, Jorge Mendez-Mendez

Keywords: task and motion planning, long-horizon, learning for planning

TL;DR: We learn a heuristic to optimize backtracking in sampling-based approaches to Task and Motion Planning.

Abstract: Autonomous robots will need the ability to make task and motion plans that involve long sequences of actions, e.g. to prepare a meal. One challenge is that the feasibility of actions late in the plan may depend on much earlier actions. This issue is exacerbated if these dependencies exist at a purely geometric level, making them difficult to express for a task planner. Backtracking is a common technique to resolve such geometric dependencies, but its time complexity limits its applicability to short-horizon dependencies. We propose an approach to account for these dependencies by learning a search heuristic for task and motion planning. We evaluate our approach on five quasi-static simulated domains and show a substantial improvement in success rate over the baselines.

Student Paper: yes

Submission Number: 437