Contact-Rich Robotic Manipulation in Construction via Zero-Shot Learning: A Diffusion Policy-Guided Adaptive Control
Industry
Source: arXiv cs.ROPublish time unverified
arXiv:2608.22100v1 Announce Type: new Abstract: Construction robotics and automation offer promising means of improving productivity, alleviating workforce shortages, and reducing workers' exposure to physically demanding tasks. However, reliable contact-rich robotic assembly remains challenging under tight tolerances, fabrication inaccuracies, and uncertain contact dynamics.