Excavation Reinforcement Learning Using Geometric Representation

TL;DR

This paper introduces a reinforcement learning approach with geometric point cloud representations to efficiently plan excavation trajectories for irregular objects in cluttered scenes, demonstrating improved real-world performance.

Contribution

It is the first to apply RL for planning excavation trajectories of irregular objects in clutter, using geometric representations to enhance training efficiency and transferability.

Findings

Representation reduces RL training time.

Representation achieves similar performance to end-to-end RL.

Representation-trained policy outperforms end-to-end RL in real scenes.

Abstract

Excavation of irregular rigid objects in clutter, such as fragmented rocks and wood blocks, is very challenging due to their complex interaction dynamics and highly variable geometries. In this paper, we adopt reinforcement learning (RL) to tackle this challenge and learn policies to plan for a sequence of excavation trajectories for irregular rigid objects, given point clouds of excavation scenes. Moreover, we separately learn a compact representation of the point cloud on geometric tasks that do not require human labeling. We show that using the representation reduces training time for RL, while achieving similar asymptotic performance compare to an end-to-end RL algorithm. When using a policy trained in simulation directly on a real scene, we show that the policy trained with the representation outperforms end-to-end RL. To our best knowledge, this paper presents the first application of RL to plan a sequence of excavation trajectories of irregular rigid objects in clutter.