Learning to unfold cloth: Scaling up world models to deformable object manipulation

TL;DR

This paper introduces an enhanced reinforcement learning approach for cloth manipulation that improves generalisation by incorporating surface normals and data augmentation, demonstrated through simulation and real-world experiments.

Contribution

The paper presents a modified DreamerV2 architecture with surface normals input and improved data handling for better cloth manipulation generalisation.

Findings

Successful in-air cloth unfolding in simulation and real robot

Enhanced generalisation across different cloth types

Zero-shot transfer to physical robot setup

Abstract

Learning to manipulate cloth is both a paradigmatic problem for robotic research and a problem of immediate relevance to a variety of applications ranging from assistive care to the service industry. The complex physics of the deformable object makes this problem of cloth manipulation nontrivial. In order to create a general manipulation strategy that addresses a variety of shapes, sizes, fold and wrinkle patterns, in addition to the usual problems of appearance variations, it becomes important to carefully consider model structure and their implications for generalisation performance. In this paper, we present an approach to in-air cloth manipulation that uses a variation of a recently proposed reinforcement learning architecture, DreamerV2. Our implementation modifies this architecture to utilise surface normals input, in addition to modiying the replay buffer and data augmentation procedures. Taken together these modifications represent an enhancement to the world model used by the robot, addressing the physical complexity of the object being manipulated by the robot. We present evaluations both in simulation and in a zero-shot deployment of the trained policies in a physical robot setup, performing in-air unfolding of a variety of different cloth types, demonstrating the generalisation benefits of our proposed architecture.