Dynamics Randomization Revisited:A Case Study for Quadrupedal Locomotion

TL;DR

This study investigates the role of dynamics randomization in sim-to-real transfer for quadrupedal robots, revealing that successful transfer can occur without it, contrary to prior assumptions, through extensive ablation and real-world testing.

Contribution

The paper provides new insights showing that dynamics randomization is not always necessary for sim-to-real transfer in quadrupedal locomotion, supported by comprehensive ablation studies and real-world experiments.

Findings

Successful sim-to-real transfer without dynamics randomization

Key design factors influencing policy robustness identified

Empirical validation across various gaits and speeds

Abstract

Understanding the gap between simulation and reality is critical for reinforcement learning with legged robots, which are largely trained in simulation. However, recent work has resulted in sometimes conflicting conclusions with regard to which factors are important for success, including the role of dynamics randomization. In this paper, we aim to provide clarity and understanding on the role of dynamics randomization in learning robust locomotion policies for the Laikago quadruped robot. Surprisingly, in contrast to prior work with the same robot model, we find that direct sim-to-real transfer is possible without dynamics randomization or on-robot adaptation schemes. We conduct extensive ablation studies in a sim-to-sim setting to understand the key issues underlying successful policy transfer, including other design decisions that can impact policy robustness. We further ground our conclusions via sim-to-real experiments with various gaits, speeds, and stepping frequencies. Additional Details: https://www.pair.toronto.edu/understanding-dr/.