Advanced Skills through Multiple Adversarial Motion Priors in Reinforcement Learning
This work addresses the challenge of reward function design for roboticists, offering a more flexible approach to learning diverse skills, though it is incremental as it builds on existing adversarial motion prior methods.
The paper tackles the problem of tedious reward tuning in reinforcement learning for robotic locomotion by extending adversarial motion priors to allow multiple, switchable motion styles, and demonstrates that skills like ducking, walking, and switching between quadrupedal and humanoid configurations can be learned simultaneously without performance loss, validated in real-world experiments with a wheeled-legged quadruped robot.
In recent years, reinforcement learning (RL) has shown outstanding performance for locomotion control of highly articulated robotic systems. Such approaches typically involve tedious reward function tuning to achieve the desired motion style. Imitation learning approaches such as adversarial motion priors aim to reduce this problem by encouraging a pre-defined motion style. In this work, we present an approach to augment the concept of adversarial motion prior-based RL to allow for multiple, discretely switchable styles. We show that multiple styles and skills can be learned simultaneously without notable performance differences, even in combination with motion data-free skills. Our approach is validated in several real-world experiments with a wheeled-legged quadruped robot showing skills learned from existing RL controllers and trajectory optimization, such as ducking and walking, and novel skills such as switching between a quadrupedal and humanoid configuration. For the latter skill, the robot is required to stand up, navigate on two wheels, and sit down. Instead of tuning the sit-down motion, we verify that a reverse playback of the stand-up movement helps the robot discover feasible sit-down behaviors and avoids tedious reward function tuning.