Robots learn to use compliance for easier part assembly
Learning to Leverage Compliance: A Policy-Admittance Learning Framework for Robotic Insertion
Robotics
Summary
In robot assembly, pushing parts together can cause problems if the robot keeps pushing even when the part is resisting. The authors designed a way for robots to learn when to push and when to yield by listening to the forces they feel during insertion. This approach helps the robot avoid wasting effort and reduces the force applied, making assembly smoother and safer. They tested it on real connector tasks and got better success and gentler handling than traditional methods.
What this means in practice
- •For industrial automation engineers: Improve robotic assembly lines with robots that adapt force application to avoid damaging parts during insertion tasks.
- •For robotic system integrators: Integrate learning-based compliance frameworks to enhance robot performance under pose and contact uncertainties in assembly setups.
Authors
Chongren Wang, Minghe Li, Honghua Dai, Zhicheng Lin, Shiyang Wei, Xiaokui Yue
Abstract
Policy learning and compliant control offer a promising route to reliable autonomous assembly under pose errors and contact uncertainty. However, combining them does not ensure coordination: the policy may continue pushing against contact while the controller yields, producing sustained loading with limited progress. To address this problem, we propose LeCo (Leverage Compliance), a policy-admittance learning framework that guides a visual policy through execution-time interaction under fixed admittance. A multirate feedback mechanism aggregates high-rate contact-interaction records into policy-transition rewards. An integrated conflict cost then characterizes sustained policy-loading/controller-unloading opposition, while a directional high-force tail cost captures continued-loading events within a transition. Together with task completion, these costs encourage the policy to leverage compliance with less unproductive loading. We evaluate LeCo on four real connector-assembly tasks, obtaining an aggregate success rate of 94%. Across tasks, mean successful-trial resultant-force and torque peaks decrease by approximately 30% and 64% relative to the comparison baseline. Reward ablation further shows that adding conflict shaping reduces median successful-trial contact-conditioned conflict density by approximately 53%. These results support learning to leverage fixed compliance by turning multirate policy-admittance interaction into complementary reward signals for effective, lower-load insertion.