HumanoidExo: Scalable Whole-Body Humanoid Manipulation via Wearable Exoskeleton

2025-10-03

AI Summary

Humanoid robot policy learning is hindered by the scarcity and high acquisition cost of high-quality demonstrations from the real world. To address this, we propose HumanoidExo: a system that captures full-body human motion via wearable exoskeletons and enables high-fidelity embodied transfer to humanoid robots through kinematic mapping and policy distillation. This approach significantly reduces the embodiment gap between human demonstration and robotic execution. Our method achieves full-body control policy training with only five real-robot demonstrations and supports learning novel skills—such as bipedal walking—entirely from exoskeleton data. We validate its effectiveness on diverse tasks including desktop manipulation, stand-squat coordination, and whole-body control. Results demonstrate substantially improved generalization to unseen environments. HumanoidExo establishes a new paradigm for low-cost, scalable humanoid policy learning, circumventing reliance on extensive real-robot data collection.

Technology Category

• Intelligent Robots: Embodied AI
• Humans and AI: Human-Aware Planning and Behavior Prediction
• Machine Learning: Imitation Learning & Inverse Reinforcement Learning

Application Category

• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• Economics, Online Markets and Human Computation: Humans versus LLMs for data annotation and labeling

Abstract

A significant bottleneck in humanoid policy learning is the acquisition of large-scale, diverse datasets, as collecting reliable real-world data remains both difficult and cost-prohibitive. To address this limitation, we introduce HumanoidExo, a novel system that transfers human motion to whole-body humanoid data. HumanoidExo offers a high-efficiency solution that minimizes the embodiment gap between the human demonstrator and the robot, thereby tackling the scarcity of whole-body humanoid data. By facilitating the collection of more voluminous and diverse datasets, our approach significantly enhances the performance of humanoid robots in dynamic, real-world scenarios. We evaluated our method across three challenging real-world tasks: table-top manipulation, manipulation integrated with stand-squat motions, and whole-body manipulation. Our results empirically demonstrate that HumanoidExo is a crucial addition to real-robot data, as it enables the humanoid policy to generalize to novel environments, learn complex whole-body control from only five real-robot demonstrations, and even acquire new skills (i.e., walking) solely from HumanoidExo data.

Problem

• Addressing humanoid data scarcity via wearable exoskeleton
• Minimizing embodiment gap between human and robot
• Enhancing humanoid generalization with diverse motion transfer

Innovation

• Transfers human motion to humanoid data
• Minimizes embodiment gap via wearable exoskeleton
• Enables learning from minimal real-robot demonstrations