Multi-Embodiment Locomotion at Scale with extreme Embodiment Randomization
This addresses the challenge of generalizable robot control for various morphologies, though it appears incremental as it builds on prior embodiment-aware architectures.
The paper tackles the problem of training a single locomotion policy to control a diverse set of 50 legged robots, achieving zero-shot transfer to unseen real-world humanoid and quadruped robots.
We present a single, general locomotion policy trained on a diverse collection of 50 legged robots. By combining an improved embodiment-aware architecture (URMAv2) with a performance-based curriculum for extreme Embodiment Randomization, our policy learns to control millions of morphological variations. Our policy achieves zero-shot transfer to unseen real-world humanoid and quadruped robots.