Scalable and General Whole-Body Control for Cross-Humanoid Locomotion

2026-02-05

AI Summary

Existing whole-body control strategies for humanoid robots are typically trained separately for specific platforms, limiting their ability to generalize across diverse morphologies. This work proposes XHugWBC, a framework that achieves zero-shot transfer of a single control policy across multiple humanoid robots for the first time. The approach internalizes a broad distribution of robot morphologies through physics-consistent morphological randomization, semantically aligned observation and action spaces, and a policy network that explicitly models the relationship between morphology and dynamics. Experiments on twelve simulated and seven real-world humanoid platforms demonstrate that the resulting universal controller exhibits strong generalization and robustness, significantly improving deployment efficiency across heterogeneous robotic systems.

Technology Category

• Intelligent Robots: Embodied AI
• Humans and AI: Human-Aware Planning and Behavior Prediction
• Philosophy and Ethics of AI: Safety, Robustness & Trustworthiness

Application Category

• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Economics, Online Markets and Human Computation: LLM based quality controls for crowd work
• User Modeling, Personalization and Recommendation: User modeling and simulation for interactive and conversational systems

Abstract

Learning-based whole-body controllers have become a key driver for humanoid robots, yet most existing approaches require robot-specific training. In this paper, we study the problem of cross-embodiment humanoid control and show that a single policy can robustly generalize across a wide range of humanoid robot designs with one-time training. We introduce XHugWBC, a novel cross-embodiment training framework that enables generalist humanoid control through: (1) physics-consistent morphological randomization, (2) semantically aligned observation and action spaces across diverse humanoid robots, and (3) effective policy architectures modeling morphological and dynamical properties. XHugWBC is not tied to any specific robot. Instead, it internalizes a broad distribution of morphological and dynamical characteristics during training. By learning motion priors from diverse randomized embodiments, the policy acquires a strong structural bias that supports zero-shot transfer to previously unseen robots. Experiments on twelve simulated humanoids and seven real-world robots demonstrate the strong generalization and robustness of the resulting universal controller.

Problem

• cross-embodiment
• whole-body control
• humanoid locomotion
• generalization
• zero-shot transfer

Innovation

• cross-embodiment
• whole-body control
• morphological randomization
• zero-shot transfer
• universal policy