URF: A Unified Robot Control-Policy Framework for Stable Contact Aware Manipulation
Published in arXiv preprint arXiv:2607.20912, 2026
URF is a framework that connects compliant action prediction with unified impedance-admittance control. Given multimodal observations, URF predicts a virtual target, a stiffness matrix, and an impedance-admittance switch ratio, which together determine when to employ admittance control for accurate motion tracking and when to fall back to impedance control for safer handling of rigid contact.
Evaluated on box-flipping and line-pressing tasks, the approach improves success rates while mitigating failure modes such as rapid force buildup, force oscillations, tool damage, and unnecessary robot safety stops.
Recommended citation: Jiyou Shin, Youngjin Seo, Jaeseog Won, Sungwon Seo, Hyunjun Kim, Seokmin Yoon, Tuan Luong, and Hyungpil Moon. (2026). "URF: A Unified Robot Control-Policy Framework for Stable Contact Aware Manipulation." arXiv preprint arXiv:2607.20912.
Download Paper