Imitation Learning for High Precision Peg-in-Hole Tasks
This work addresses the challenge of enabling industrial robots to perform high-precision, contact-rich tasks with greater speed and accuracy, which is a problem for automation in manufacturing.
This paper tackles the problem of high-precision peg-in-hole insertion using industrial robot manipulators. They demonstrate that generative adversarial imitation learning (GAIL) can successfully achieve this task with 10 µm and 6 µm peg-hole clearances, reducing insertion time from over 20 seconds to under 15 seconds.
Industrial robot manipulators are not able to match the precision and speed with which humans are able to execute contact rich tasks even to this day. Therefore, as a means overcome this gap, we demonstrate generative methods for imitating a peg-in-hole insertion task in a 6-DOF robot manipulator. In particular, generative adversarial imitation learning (GAIL) is used to successfully achieve this task with a 10 um, and a 6 um peg-hole clearance on the Yaskawa GP8 industrial robot. Experimental results show that the policy successfully learns within 20 episodes from a handful of human expert demonstrations on the robot (i.e., < 10 tele-operated robot demonstrations). The insertion time improves from > 20 seconds (which also includes failed insertions) to < 15 seconds, thereby validating the effectiveness of this approach.