Controlling the Solo12 quadruped robot with deep reinforcement learning

Quadruped robots require robust and general locomotion skills to exploit their mobility potential in complex and challenging environments. In this work, we present an implementation of a robust end-to-end learning-based controller on the Solo12 quadruped. Our method is based on deep reinforcement learning of joint impedance references. The resulting control policies follow a commanded velocity reference while being efficient in its energy consumption and easy to deploy. We detail the learning procedure and method for transfer on the real robot. We show elaborate experiments. Finally, we present experimental results of the learned locomotion on various grounds indoors and outdoors. These results show that the Solo12 robot is a suitable open-source platform for research combining learning and control because of the easiness in transferring and deploying learned controllers.

Automated summary

This article presents an implementation of an end-to-end learning-based controller on the Solo12 quadruped robot, which is achieved by using deep reinforcement learning to learn joint impedance references. The resulting control policies are able to efficiently follow commanded velocity references, while being energy-efficient and easy to deploy.