Robot adapts grip strength for safer part insertion tasks
Constraint-Grounded Reinforcement Learning for Variable Impedance Control in Contact-Rich Robotic Insertion
Robotics
Summary
Robots often need to push parts together without pressing too hard, but the right pushing force can change depending on the task. The authors created a method called Constraint-Grounded Reinforcement Learning (CG-RL) that helps robots adjust how hard they push automatically while inserting parts. This way, the robot avoids damage or failure without anyone having to set the pushing force by hand. Their tests showed the robot succeeded much more often with this method than using a fixed pushing strength.
What this means in practice
- •For industrial robot operators: Enable robots to automatically adjust pushing strength during part assembly tasks to reduce manual controller tuning and improve success rates.
- •For robotics system integrators: Incorporate variable impedance control policies that adapt to different force constraints online without retraining for diverse insertion scenarios.
Authors
Lin He, Min Deng
Abstract
In robotic insertion under uncertain contact, the axial force limit and the appropriate controller gain vary across tasks. As a result, a single fixed gain is unlikely to remain suitable across different task conditions, making conventional impedance controllers reliant on manual retuning. To eliminate manual retuning, we propose Constraint-Grounded Reinforcement Learning (CG-RL), a variable impedance framework for online gain adaptation. Conditioned on the force limit and contact feedback, the policy outputs a residual motion, an insertion rate, and a requested gain. The controller projects this gain into the admissible range without exposing the range itself to the policy. This separation allows a single policy to operate under different force limits without retraining or manual retuning. We evaluate CG-RL on simulated oblique insertion across five training seeds. CG-RL achieves an $85.8\pm7.7\%$ (mean $\pm$ SD) success rate of insertions without violating the force limit, while keeping the applied gain within the admissible range. As a comparison, a fixed-gain baseline using the midpoint gain achieves a success rate of $50.1\%$. The policy adapts its insertion rate continuously to the specified force limit and further generalizes to more permissive force limits above the training range. In contrast, the same actor without force-limit input does not exhibit this adaptation. The applied gain is guaranteed to remain within the admissible range, while force-limit satisfaction is validated empirically rather than guaranteed formally.