Robust In-Hand Manipulation via Priors in Reinforcement Learning and Mechanical Design

2026-07-13

AI Summary

This work addresses the challenge of stable dexterous manipulation under conditions lacking external sensing, where contact uncertainties and gravitational disturbances often compromise grasp stability. The authors propose a novel approach that synergistically integrates reinforcement learning with mechanical design: a global grasp quality prior derived from classical grasp analysis is incorporated into the reward function, while fingertip curvature geometry is engineered to optimize local contact interfaces. This co-design strategy uniquely embeds two complementary physical priors—global grasp stability and local contact mechanics—into both the learning framework and hardware morphology. Experimental results demonstrate significant improvements in rotational efficiency, grasp robustness, and disturbance rejection across four palm orientations and three object types, with enhanced sim-to-real transferability for rolling manipulation tasks.

Abstract

In-hand manipulation without external sensing is challenging due to uncertainties from finger-object contacts and disturbances by gravity. While reinforcement learning has shown promise in learning complex finger gaiting, existing approaches do not prioritize maintaining well-conditioned grasps for sustained manipulation. We introduce two complementary physics priors for robust in-hand rolling: a global grasp-quality prior derived from classical grasp analysis and a local contact-geometry prior based on fingertip curvature. The grasp-quality prior is used as a dense reward-shaping term that encourages well-distributed contacts with improved worst-case wrench resistance. The contact-geometry prior is expressed in the fingertip geometry that mechanically shapes the contact interface toward task-aligned rolling while reducing off-axis drift. We evaluate the effect of these priors on learning in-hand rolling manipulation for a multifingered robotic hand manipulating three different objects at four palm orientations. Results show significant improvement in rotation efficiency, grasp stability, and disturbance rejection, suggesting that physics priors embedded in both learning and fingertip morphology improve task robustness and sim-to-real transfer. An overview video can be found at https://youtu.be/pdd1wHxQnJM?si=dM-U5kiiPTYsk3Pk.

Problem

• in-hand manipulation
• grasp stability
• contact uncertainty
• disturbance rejection
• reinforcement learning

Innovation

• physics priors
• in-hand manipulation
• grasp quality
• contact geometry
• reinforcement learning