Precise SE(3) End-Effector Tracking in Whole-Body Humanoid Control

Key points

• Precise end-effector tracking during humanoid whole-body motion is challenging due to floating-base oscillations, gravity, dynamic coupling, and locomotion-induced disturbances.1
• GAC provides structured $\SE$ task-space feedback and generates nominal arm joint-position targets, while residual RL compensates for unmodeled dynamics and coordinates locomotion and balance in the shared joint-position action space.1
• ResGAC is validated on a real Unitree G1 humanoid.1
• ResGAC achieves $90%$ success in a standing peg-in-hole task compared with $50%$ for SONIC, and accurate world-frame $\SE$ end-effector pose tracking during lower-body motion.1

Sources (1)

• [1]Precise SE(3) End-Effector Tracking in Whole-Body Humanoid Control
  arXiv (AI, ML, NLP, CV, robotics, multi-agent) · Oct 7, 05:34 AM

  We propose ResGAC, a whole-body humanoid controller for precise end-effector pose tracking that combines geometric admittance control (GAC) with residual reinforcement learning.

  Precise end-effector tracking during humanoid whole-body motion is challenging due to floating-base oscillations, gravity, dynamic coupling, and locomotion-induced disturbances.

Extractive summary: sentences quoted from the sources.