HULK: Learning Whole-Body Forceful Loco-Manipulation for Humanoids

What happened

arXiv:2610.08970v1 Announce Type: new Abstract: Humanoid loco-manipulation of large, heavy objects demands forceful interaction across the entire body. A capture-point control barrier function augments the wrist-force teacher during training to improve balance under load.

However, such payloads shift a humanoid's center of mass and impose sustained loads across the upper body, challenging balance and command tracking. Evaluation spans simulation and the Unitree G1. In simulation, the teacher with the barrier function achieves the lowest forward and lateral velocity tracking errors at 10 kg per arm among evaluated controllers and reduces aggregate divergent component of motion (DCM) excursion magnitude by 35.7% relative to MPC-guided reinforcement learning alone.