Track papers across labs, topics, and trends.
Search papers, labs, and topics across Lattice.
This paper introduces the Unified Robot Control-Policy Framework (URF), which integrates compliant action prediction with impedance-admittance control to enhance robot manipulation in contact-rich environments. By predicting a virtual target, a stiffness matrix, and a switch ratio for controller behavior, URF effectively manages the transition between accurate motion tracking and safe rigid contact. Experimental results demonstrate that URF significantly improves task success rates and reduces failure modes compared to traditional admittance-only execution in box-flipping and line-pressing tasks.
URF not only predicts compliant actions but also optimizes controller behavior, leading to safer and more effective robot manipulation in contact-heavy scenarios.
Learning-based manipulation policies usually predict robot actions from sensory observations and leave their execution to a separate low-level controller. In rigid contact, this separation can be problematic: the same motion to a virtual target or compliant motion command can lead to unstable contact, tracking error, excessive loading, or tool damage, depending on the low-level controller. In this paper, we propose a \textit{Unified Robot Control-Policy Framework} (URF), which 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. The switch ratio determines when the controller should behave more like admittance control for accurate motion tracking and when it should move toward impedance control for safer rigid contact. Because demonstration data do not provide ground-truth environment stiffness, we construct switch-ratio labels from measured contact forces and use them to supervise controller-mode prediction. Across box-flipping and line-pressing tasks, URF achieves higher task success rates while reducing failure modes observed with admittance-only execution, including rapid force buildup, large force oscillations, tool breakage, and robot safety stops. These results suggest that contact-aware policies benefit from predicting not only compliant actions but also the controller behavior used to execute them. Project page: https://jiyou384.github.io/urf_project_page/