BiAssemble: Learning Collaborative Affordance for Bimanual Geometric Assembly

TL;DR

This paper introduces BiAssemble, a method that enables robots to learn and generalize geometric affordances for bimanual assembly tasks, demonstrated on a new real-world benchmark with diverse parts.

Contribution

We propose a novel affordance learning framework for bimanual geometric assembly that generalizes across diverse shapes and includes a real-world benchmark for evaluation.

Findings

Our approach outperforms previous affordance-based methods.

The method demonstrates effective generalization to varied geometric fragments.

Extensive experiments validate the superiority of our approach.

Abstract

Shape assembly, the process of combining parts into a complete whole, is a crucial robotic skill with broad real-world applications. Among various assembly tasks, geometric assembly--where broken parts are reassembled into their original form (e.g., reconstructing a shattered bowl)--is particularly challenging. This requires the robot to recognize geometric cues for grasping, assembly, and subsequent bimanual collaborative manipulation on varied fragments. In this paper, we exploit the geometric generalization of point-level affordance, learning affordance aware of bimanual collaboration in geometric assembly with long-horizon action sequences. To address the evaluation ambiguity caused by geometry diversity of broken parts, we introduce a real-world benchmark featuring geometric variety and global reproducibility. Extensive experiments demonstrate the superiority of our approach over both previous affordance-based and imitation-based methods. Project page: https://sites.google.com/view/biassembly/.