Stiffness Copilot: An Impedance Policy for Contact-Rich Teleoperation
This addresses the problem of balancing safety and efficiency in robot teleoperation for manipulation tasks, representing an incremental improvement with a novel method for a known bottleneck.
The paper tackled the challenge of selecting robot impedance in contact-rich teleoperation by introducing Stiffness Copilot, a vision-based policy that adjusts impedance online, achieving safety comparable to low stiffness and efficiency matching high stiffness in a human-subject study.
In teleoperation of contact-rich manipulation tasks, selecting robot impedance is critical but difficult. The robot must be compliant to avoid damaging the environment, but stiff to remain responsive and to apply force when needed. In this paper, we present Stiffness Copilot, a vision-based policy for shared-control teleoperation in which the operator commands robot pose and the policy adjusts robot impedance online. To train Stiffness Copilot, we first infer direction-dependent stiffness matrices in simulation using privileged contact information. We then use these matrices to supervise a lightweight vision policy that predicts robot stiffness from wrist-camera images and transfers zero-shot to real images at runtime. In a human-subject study, Stiffness Copilot achieved safety comparable to using a constant low stiffness while matching the efficiency of using a constant high stiffness.