Learning a Contact-Adaptive Controller for Robust, Efficient Legged Locomotion

TL;DR

This paper introduces a hierarchical control framework combining model-based control and reinforcement learning to create a contact-adaptive quadruped controller that is robust, energy-efficient, and capable of handling unseen environments.

Contribution

A novel hierarchical framework that integrates RL with model-based control for adaptive, robust quadruped locomotion without requiring online adaptation.

Findings

Controller is up to 85% more energy efficient.

Demonstrates robustness in unseen environments.

Successfully deployed on a physical robot without additional adaptation.

Abstract

We present a hierarchical framework that combines model-based control and reinforcement learning (RL) to synthesize robust controllers for a quadruped (the Unitree Laikago). The system consists of a high-level controller that learns to choose from a set of primitives in response to changes in the environment and a low-level controller that utilizes an established control method to robustly execute the primitives. Our framework learns a controller that can adapt to challenging environmental changes on the fly, including novel scenarios not seen during training. The learned controller is up to 85~percent more energy efficient and is more robust compared to baseline methods. We also deploy the controller on a physical robot without any randomization or adaptation scheme.