Contact-conditioned learning of multi-gait locomotion policies

TL;DR

This paper investigates how goal representation affects multi-gait policy learning in legged robots, showing that contact-conditioned policies improve generalization and versatility across different gaits.

Contribution

It introduces contact-conditioned goal representation for multi-gait policies, demonstrating superior generalization over other methods in simulation.

Findings

Contact-conditioned policies outperform other goal representations.

Policies generalize better outside training distribution.

Effective for both bipedal and quadrupedal robots.

Abstract

In this paper, we examine the effects of goal representation on the performance and generalization in multi-gait policy learning settings for legged robots. To study this problem in isolation, we cast the policy learning problem as imitating model predictive controllers that can generate multiple gaits. We hypothesize that conditioning a learned policy on future contact switches is a suitable goal representation for learning a single policy that can generate a variety of gaits. Our rationale is that policies conditioned on contact information can leverage the shared structure between different gaits. Our extensive simulation results demonstrate the validity of our hypothesis for learning multiple gaits on a bipedal and a quadrupedal robot. Most interestingly, our results show that contact-conditioned policies generalize much better than other common goal representations in the literature, when the robot is tested outside the distribution of the training data.