Exciting Contact Modes in Differentiable Simulations for Robot Learning

TL;DR

This paper introduces an information-theoretic approach to actively excite contact modes in differentiable simulators, improving robot parameter estimation and reducing the sim-to-real gap.

Contribution

It presents a novel optimal experimental design method for contact mode excitation in differentiable simulation, enhancing parameter identification accuracy.

Findings

Achieved at least 84% error reduction in parameter estimation

Demonstrated higher information gains compared to random sampling

Improved contact mode excitation in robot simulations

Abstract

In this paper, we explore an approach to actively plan and excite contact modes in differentiable simulators as a means to tighten the sim-to-real gap. We propose an optimal experimental design approach derived from information-theoretic methods to identify and search for information-rich contact modes through the use of contact-implicit optimization. We demonstrate our approach on a robot parameter estimation problem with unknown inertial and kinematic parameters which actively seeks contacts with a nearby surface. We show that our approach improves the identification of unknown parameter estimates over experimental runs by an estimate error reduction of at least $\sim 84\%$ when compared to a random sampling baseline, with significantly higher information gains.