Search papers, labs, and topics across Lattice.
This paper introduces an adaptive stiffness control framework for robots engaged in physical human-robot collaboration, utilizing generative action-chunk sampling based on RGB images and joint-torque estimates. The method allows robots to dynamically adjust their stiffness and damping in response to the variability of human actions, enhancing compliance when multiple future motions are possible. Experimental results demonstrate that this approach significantly improves task success rates in collaborative transport scenarios, achieving an average of 0.95 compared to 0.83 for fixed-stiffness and 0.69 for deterministic methods.
Variation in sampled actions can be leveraged as a real-time control signal to optimize robot compliance and assistance in human-robot collaboration.
Physical human-robot collaboration requires a robot to provide assistance when human intention is clear while remaining compliant when several future motions are plausible. We present an adaptive stiffness framework based on generative action-chunk sampling. Conditioned on an RGB image and external joint-torque estimates, the policy samples multiple future action chunks from an observation-conditioned prior. Variation among the sampled action chunks is used to continuously adapt joint stiffness and damping. Greater variation makes the robot more compliant to facilitate human guidance, whereas lower variation provides firmer assistance. In a real-world collaborative transport task with four possible directions, the proposed method achieved an average success rate of 0.95, compared with 0.83 for a fixed-stiffness ablation and 0.69 for a deterministic baseline. Near direction determination, variation among the sampled action chunks increased and the controller accordingly reduced stiffness. These results suggest that variation among actions sampled by a generative policy can serve as an online control signal for balancing assistance and compliance in physical human-robot interaction.
