On Surprising Effects of Risk-Aware Domain Randomization for Contact-Rich Sampling-based Predictive Control

2026-05-04

AI Summary

This work addresses the challenge of degraded trajectory replay quality in contact-rich model-based predictive control due to model uncertainty, a setting where the potential of domain randomization remains underexplored. The authors propose a risk-aware domain randomization framework that integrates three replay aggregation strategies—average, optimistic, and pessimistic—and systematically analyze their impact on the cost landscape perceived by the optimizer in the Push-T task. Experimental results demonstrate that the choice of risk-sensitive aggregation significantly alters controller performance, enhancing robustness to model errors while reshaping the cost structure of attraction basins associated with contact actions. These findings offer a novel perspective on risk-aware control in highly interactive manipulation tasks.

Abstract

Domain randomization (DR) is widely used in policy learning to improve robustness to modeling error, but remains underexplored in contact-rich sampling-based predictive control (SPC), where rollout quality is highly sensitive to uncertainty. In this work, we take the first step by studying risk-aware DR in predictive sampling on a simple yet representative Push-T task, comparing average, optimistic, and pessimistic rollout aggregations under randomized model instances. Our initial results suggest that DR affects not only robustness to model error, but also the effective cost landscape seen by the sampling-based optimizer, by reshaping the basin of attraction around contact-producing actions. This opens up potential for exploring better grounded risk-aware contact-rich SPC under model uncertainty. Video: https://youtu.be/f1F0ALXxhSM

Problem

• domain randomization
• sampling-based predictive control
• contact-rich
• model uncertainty
• risk-aware

Innovation

• risk-aware domain randomization
• sampling-based predictive control
• contact-rich manipulation
• model uncertainty
• rollout aggregation