Long-Horizon Prediction and Uncertainty Propagation with Residual Point Contact Learners

TL;DR

This paper introduces a self-supervised residual learning approach to improve rigid-body contact simulation accuracy and uncertainty propagation, demonstrated through planar dice roll predictions.

Contribution

It presents a novel residual contact model that refines simulator predictions and propagates uncertainty, enhancing simulation fidelity for robotic contact tasks.

Findings

Outperforms state-of-the-art contact prediction methods

Effectively propagates uncertainty in contact simulations

Improves accuracy in planar dice roll outcome predictions

Abstract

The ability to simulate and predict the outcome of contacts is paramount to the successful execution of many robotic tasks. Simulators are powerful tools for the design of robots and their behaviors, yet the discrepancy between their predictions and observed data limit their usability. In this paper, we propose a self-supervised approach to learning residual models for rigid-body simulators that exploits corrections of contact models to refine predictive performance and propagate uncertainty. We empirically evaluate the framework by predicting the outcomes of planar dice rolls and compare it's performance to state-of-the-art techniques.