NuExo: A Wearable Exoskeleton Covering all Upper Limb ROM for Outdoor Data Collection and Teleoperation of Humanoid Robots

TL;DR

NuExo introduces a lightweight backpack-mounted exoskeleton with a novel shoulder mechanism that achieves full upper-limb ROM while minimizing actuators, enabling ergonomic outdoor use. It pairs this hardware with a unified, calibration-free teleoperation framework and a comprehensive multimodal data-collection system (including force-torque, joint positions, and egocentric vision) that supports multiple humanoid platforms. Empirical results demonstrate superior ROM fidelity and robustness to dynamic perturbations compared with inertial motion capture, high teleoperation accuracy (average around $0.015$ rad; peaks $0.05$–$0.08$ rad), and rich datasets suitable for imitation learning. The work offers a practical, versatile path for data collection and skill transfer in humanoid robotics, reducing training requirements and enabling outdoor deployment.

Abstract

The evolution from motion capture and teleoperation to robot skill learning has emerged as a hotspot and critical pathway for advancing embodied intelligence. However, existing systems still face a persistent gap in simultaneously achieving four objectives: accurate tracking of full upper limb movements over extended durations (Accuracy), ergonomic adaptation to human biomechanics (Comfort), versatile data collection (e.g., force data) and compatibility with humanoid robots (Versatility), and lightweight design for outdoor daily use (Convenience). We present a wearable exoskeleton system, incorporating user-friendly immersive teleoperation and multi-modal sensing collection to bridge this gap. Due to the features of a novel shoulder mechanism with synchronized linkage and timing belt transmission, this system can adapt well to compound shoulder movements and replicate 100% coverage of natural upper limb motion ranges. Weighing 5.2 kg, NuExo supports backpack-type use and can be conveniently applied in daily outdoor scenarios. Furthermore, we develop a unified intuitive teleoperation framework and a comprehensive data collection system integrating multi-modal sensing for various humanoid robots. Experiments across distinct humanoid platforms and different users validate our exoskeleton's superiority in motion range and flexibility, while confirming its stability in data collection and teleoperation accuracy in dynamic scenarios.

Figures (10)

• Figure 1: NuExo: A backpack-mounted active-joint humanoid robot exoskeleton teleoperation system with large motion range.
• Figure 2: Overview of the NuExo teleoperation exoskeleton data collection system (a) and the NuExo in daily scene (b).
• Figure 3: The schematic representation of the shoulder mechanical structure of the exoskeleton tracking the dynamically changing center of GH joint in the horizontal plane.
• Figure 4: The schematic illustration of how the mechanical structure tracks the dynamically changing center of GH joint in the vertical direction.
• Figure 5: Control diagram of the Teleoperation. the shoulder and wrist posture (the elbow and hand joints position) of the human arm with the exoskeleton $\boldsymbol{q}^{m}$ and that of the humanoid robot $\boldsymbol{q}^{s}$, and the angular velocity and joints rotate of the both ${\hat{\dot{\boldsymbol{q}}}}^{m}$, ${\hat{\dot{\boldsymbol{q}}}}^{s}$. They are transmitted to the teleoperation controller. The differences between $\boldsymbol{q}^{m}$, $\boldsymbol{q}^{s}$ and ${\hat{\dot{\boldsymbol{q}}}}^{m}$,${\hat{\dot{\boldsymbol{q}}}}^{s}$ form the control target input $\boldsymbol{q}_{t}$, ${\dot{\boldsymbol{q}}}_{t}$ to the impedance controller, which outputs the torque control varibles ${\boldsymbol{\tau}}_{cmd}^{s}$ for each component of the controlled robot. Regarding Exo. control, we implement our previous work the binding alignment strategy (BAS) and the full-arm coordination mechanism (FCM) cheng2025flexibleexoskeletoncontrolbasedto produce the coordination control variable $\boldsymbol{\tau}_{h}^{m}$Cheng2024tmech, which along with the dynamics compensation $\boldsymbol{\tau}_{com}$ (consist of inertia $\hat{\boldsymbol{M}}$, Coriolis force $\hat{\boldsymbol{h}}$, gravity $\hat{\boldsymbol{g}}$, and friction compensation $\hat{\boldsymbol{f}}$), constitutes the control variable of the exoskeleton motors $\boldsymbol{\tau}_{cmd}^{m}$.