Towards Embodiment Scaling Laws in Robot Locomotion

Bo Ai, Liu Dai, Nico Bohlinger, Dichen Li, Tongzhou Mu, Zhanxin Wu, K. Fay, Henrik I Christensen, Jan Peters, Hao Su
Proceedings of The 9th Conference on Robot Learning, PMLR 305:3483-3515, 2025.

Abstract

Developing generalist agents that operate across diverse tasks, environments, and robot embodiments is a grand challenge in robotics and artificial intelligence. While substantial progress has been made in cross-task and cross-environment generalization, achieving broad generalization to novel embodiments remains elusive. In this work, we study embodiment scaling laws — the hypothesis that increasing the quantity of training embodiments improves generalization to unseen ones. To explore this, we procedurally generate a dataset of $\sim$1,000 varied robot embodiments, spanning humanoids, quadrupeds, and hexapods, and train embodiment-specific reinforcement learning experts for legged locomotion. We then distill these experts into a single generalist policy capable of handling diverse observation and action spaces. Our large-scale study reveals that generalization performance improves with the number of training embodiments. Notably, a policy trained on the full dataset zero-shot transfers to diverse unseen embodiments in both simulation and real-world evaluations. These results provide preliminary empirical evidence for embodiment scaling laws and suggest that scaling up embodiment quantity may serve as a foundation for building generalist robot agents.

Cite this Paper

BibTeX
@InProceedings{pmlr-v305-ai25a, title = {Towards Embodiment Scaling Laws in Robot Locomotion}, author = {Ai, Bo and Dai, Liu and Bohlinger, Nico and Li, Dichen and Mu, Tongzhou and Wu, Zhanxin and Fay, K. and Christensen, Henrik I and Peters, Jan and Su, Hao}, booktitle = {Proceedings of The 9th Conference on Robot Learning}, pages = {3483--3515}, year = {2025}, editor = {Lim, Joseph and Song, Shuran and Park, Hae-Won}, volume = {305}, series = {Proceedings of Machine Learning Research}, month = {27--30 Sep}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v305/main/assets/ai25a/ai25a.pdf}, url = {https://proceedings.mlr.press/v305/ai25a.html}, abstract = {Developing generalist agents that operate across diverse tasks, environments, and robot embodiments is a grand challenge in robotics and artificial intelligence. While substantial progress has been made in cross-task and cross-environment generalization, achieving broad generalization to novel embodiments remains elusive. In this work, we study embodiment scaling laws — the hypothesis that increasing the quantity of training embodiments improves generalization to unseen ones. To explore this, we procedurally generate a dataset of $\sim$1,000 varied robot embodiments, spanning humanoids, quadrupeds, and hexapods, and train embodiment-specific reinforcement learning experts for legged locomotion. We then distill these experts into a single generalist policy capable of handling diverse observation and action spaces. Our large-scale study reveals that generalization performance improves with the number of training embodiments. Notably, a policy trained on the full dataset zero-shot transfers to diverse unseen embodiments in both simulation and real-world evaluations. These results provide preliminary empirical evidence for embodiment scaling laws and suggest that scaling up embodiment quantity may serve as a foundation for building generalist robot agents.} }
Endnote
%0 Conference Paper %T Towards Embodiment Scaling Laws in Robot Locomotion %A Bo Ai %A Liu Dai %A Nico Bohlinger %A Dichen Li %A Tongzhou Mu %A Zhanxin Wu %A K. Fay %A Henrik I Christensen %A Jan Peters %A Hao Su %B Proceedings of The 9th Conference on Robot Learning %C Proceedings of Machine Learning Research %D 2025 %E Joseph Lim %E Shuran Song %E Hae-Won Park %F pmlr-v305-ai25a %I PMLR %P 3483--3515 %U https://proceedings.mlr.press/v305/ai25a.html %V 305 %X Developing generalist agents that operate across diverse tasks, environments, and robot embodiments is a grand challenge in robotics and artificial intelligence. While substantial progress has been made in cross-task and cross-environment generalization, achieving broad generalization to novel embodiments remains elusive. In this work, we study embodiment scaling laws — the hypothesis that increasing the quantity of training embodiments improves generalization to unseen ones. To explore this, we procedurally generate a dataset of $\sim$1,000 varied robot embodiments, spanning humanoids, quadrupeds, and hexapods, and train embodiment-specific reinforcement learning experts for legged locomotion. We then distill these experts into a single generalist policy capable of handling diverse observation and action spaces. Our large-scale study reveals that generalization performance improves with the number of training embodiments. Notably, a policy trained on the full dataset zero-shot transfers to diverse unseen embodiments in both simulation and real-world evaluations. These results provide preliminary empirical evidence for embodiment scaling laws and suggest that scaling up embodiment quantity may serve as a foundation for building generalist robot agents.
APA
Ai, B., Dai, L., Bohlinger, N., Li, D., Mu, T., Wu, Z., Fay, K., Christensen, H.I., Peters, J. & Su, H.. (2025). Towards Embodiment Scaling Laws in Robot Locomotion. Proceedings of The 9th Conference on Robot Learning, in Proceedings of Machine Learning Research 305:3483-3515 Available from https://proceedings.mlr.press/v305/ai25a.html.

Related Material