Automating Deformable Gasket Assembly

TL;DR

This paper compares four methods, including a deep imitation learning policy and procedural algorithms, for automating the precise and long-horizon task of deformable gasket assembly, demonstrating varying success rates in physical trials.

Contribution

It introduces and evaluates four different approaches for gasket assembly, including a deep learning policy trained on human demonstrations, providing insights into their effectiveness.

Findings

Binary+ algorithm succeeds in all straight channel trials

Deep imitation policy achieves high success but is slower

Procedural algorithms vary in success rates

Abstract

In Gasket Assembly, a deformable gasket must be aligned and pressed into a narrow channel. This task is common for sealing surfaces in the manufacturing of automobiles, appliances, electronics, and other products. Gasket Assembly is a long-horizon, high-precision task and the gasket must align with the channel and be fully pressed in to achieve a secure fit. To compare approaches, we present 4 methods for Gasket Assembly: one policy from deep imitation learning and three procedural algorithms. We evaluate these methods with 100 physical trials. Results suggest that the Binary+ algorithm succeeds in 10/10 on the straight channel whereas the learned policy based on 250 human teleoperated demonstrations succeeds in 8/10 trials and is significantly slower. Code, CAD models, videos, and data can be found at https://berkeleyautomation.github.io/robot-gasket/