You're Pushing My Buttons: Instrumented Learning of Gentle Button Presses

TL;DR

This paper investigates using sensorized contact information during training to improve contact-rich manipulation policies, demonstrating that auxiliary contact signals can reduce contact force without affecting success rates.

Contribution

It introduces a method of sensorizing objects with microphones and privileged signals to enhance contact detection and force regulation in manipulation policies.

Findings

Instrumentation-guided audio representations reduce contact force.

Success rates are similar across methods with and without instrumentation.

Sensorized training improves contact safety without sacrificing performance.

Abstract

Learning contact-rich manipulation is difficult from cameras and proprioception alone because contact events are only partially observed. We test whether training-time instrumentation, i.e., object sensorisation, can improve policy performance without creating deployment-time dependencies. Specifically, we study button pressing as a testbed and use a microphone fingertip to capture contact-relevant audio. We use an instrumented button-state signal as privileged supervision to fine-tune an audio encoder into a contact event detector. We combine the resulting representation with imitation learning using three strategies, such that the policy only uses vision and audio during inference. Button press success rates are similar across methods, but instrumentation-guided audio representations consistently reduce contact force. These results support instrumentation as a practical training-time auxiliary objective for learning contact-rich manipulation policies.