Reinforcement Learning on Variable Impedance Controller for High-Precision Robotic Assembly

TL;DR

This paper presents a reinforcement learning approach that integrates variable impedance control and force/torque feedback for high-precision robotic assembly tasks, demonstrating improved adaptability and accuracy.

Contribution

It introduces a novel RL framework combining force control with neural network generalization for precise robotic assembly, inspired by human force perception.

Findings

Effective force/torque integration improves assembly precision

Neural network generalizes to environmental variations

Achieves high success rate in Siemens Robot Learning Challenge

Abstract

Precise robotic manipulation skills are desirable in many industrial settings, reinforcement learning (RL) methods hold the promise of acquiring these skills autonomously. In this paper, we explicitly consider incorporating operational space force/torque information into reinforcement learning; this is motivated by humans heuristically mapping perceived forces to control actions, which results in completing high-precision tasks in a fairly easy manner. Our approach combines RL with force/torque information by incorporating a proper operational space force controller; where we also exploit different ablations on processing this information. Moreover, we propose a neural network architecture that generalizes to reasonable variations of the environment. We evaluate our method on the open-source Siemens Robot Learning Challenge, which requires precise and delicate force-controlled behavior to assemble a tight-fit gear wheel set.