Thor: Towards Human-Level Whole-Body Reactions for Intense Contact-Rich Environments

TL;DR

Thor is a novel humanoid framework that enables human-like, adaptive whole-body reactions in contact-rich environments, significantly improving force-interaction performance through a force-adaptive reward and a decoupled reinforcement learning architecture.

Contribution

The paper introduces Thor, a new reinforcement learning-based framework with a force-adaptive reward and decoupled control architecture for humanoids in contact-rich tasks, demonstrating superior force-interaction capabilities.

Findings

Achieves a peak pulling force of 167.7 N, 68.9% higher than baselines.

Capable of pulling a loaded rack and opening a fire door with one hand.

Substantially outperforms baseline methods in force-interaction tasks.

Abstract

Humanoids hold great potential for service, industrial, and rescue applications, in which robots must sustain whole-body stability while performing intense, contact-rich interactions with the environment. However, enabling humanoids to generate human-like, adaptive responses under such conditions remains a major challenge. To address this, we propose Thor, a humanoid framework for human-level whole-body reactions in contact-rich environments. Based on the robot's force analysis, we design a force-adaptive torso-tilt (FAT2) reward function to encourage humanoids to exhibit human-like responses during force-interaction tasks. To mitigate the high-dimensional challenges of humanoid control, Thor introduces a reinforcement learning architecture that decouples the upper body, waist, and lower body. Each component shares global observations of the whole body and jointly updates its parameters. Finally, we deploy Thor on the Unitree G1, and it substantially outperforms baselines in force-interaction tasks. Specifically, the robot achieves a peak pulling force of 167.7 N (approximately 48% of the G1's body weight) when moving backward and 145.5 N when moving forward, representing improvements of 68.9% and 74.7%, respectively, compared with the best-performing baseline. Moreover, Thor is capable of pulling a loaded rack (130 N) and opening a fire door with one hand (60 N). These results highlight Thor's effectiveness in enhancing humanoid force-interaction capabilities.