AMO :
Adaptive Motion Optimization for Hyper-Dexterous Humanoid Whole-Body Control
Real-World AMO Controller in Action
Humanoid robots executing dexterous manipulation with expanded workspace via a teleoperation interface. (Played in 3x)
AMO Teleoperation Interface
We use multi-task IK to map 3 poses from a VR device to 18-dim commands (7 for each arm, roll, pitch, yaw of torso orientation, and base height) for the AMO controller to execute. (Played in 3x)
Our teleoperation interface provides intuitive whole-body teleop for high-DoF robot control
AMO performance in MuJoCo
Autonomous Policies
Abstract
Humanoid robots derive much of their dexterity from hyper-dexterous whole-body movements, enabling tasks that require a large operational workspace—such as picking objects off the ground. However, achieving these capabilities on real humanoids remains challenging due to their high degrees of freedom (DoF) and nonlinear dynamics. We propose Adaptive Motion Optimization (AMO), a framework that integrates sim-to-real reinforcement learning (RL) with trajectory optimization for real-time, adaptive whole-body control. To mitigate distribution bias in motion imitation RL, we construct a hybrid AMO dataset and train a network capable of robust, on-demand adaptation to potentially O.O.D. commands. We validate AMO in simulation and on a 29-DoF Unitree G1 humanoid robot, demonstrating superior stability and an expanded workspace compared to strong baselines. Finally, we show that AMO's consistent performance supports autonomous task execution via imitation learning, underscoring the system's versatility and robustness.
We thank Jiajian Fu for designing Active Head 2.0.
BibTeX
@article{li2025amo,
title={AMO: Adaptive Motion Optimization for Hyper-Dexterous Humanoid Whole-Body Control},
author={Li, Jialong and Cheng, Xuxin and Huang, Tianshu and Yang, Shiqi and Qiu, Rizhao and Wang, Xiaolong},
journal={Robotics: Science and Systems 2025},
year={2025}
}