Learning to Reorient Objects with Stable Placements Afforded by Extrinsic Supports

TL;DR

This paper introduces a pipeline for predicting stable object placements from point clouds to facilitate reorientation tasks, demonstrating improved accuracy and feasible manipulation sequences in both simulated and real-world settings.

Contribution

It presents a novel multi-stage pipeline and an algorithm for constructing manipulation graphs, enabling effective object reorientation with enhanced placement accuracy.

Findings

20% improvement in placement accuracy over baseline

Effective generalization to novel objects in simulation

Feasible manipulation sequences identified in experiments

Abstract

Reorienting objects by using supports is a practical yet challenging manipulation task. Owing to the intricate geometry of objects and the constrained feasible motions of the robot, multiple manipulation steps are required for object reorientation. In this work, we propose a pipeline for predicting various object placements from point clouds. This pipeline comprises three stages: a pose generation stage, followed by a pose refinement stage, and culminating in a placement classification stage. We also propose an algorithm to construct manipulation graphs based on point clouds. Feasible manipulation sequences are determined for the robot to transfer object placements. Both simulated and real-world experiments demonstrate that our approach is effective. The simulation results underscore our pipeline's capacity to generalize to novel objects in random start poses. Our predicted placements exhibit a 20% enhancement in accuracy compared to the state-of-the-art baseline. Furthermore, the robot finds feasible sequential steps in the manipulation graphs constructed by our algorithm to accomplish object reorientation manipulation.