Reproduce “CW-WAM: A Structured Contact-Wrench World Action Model for Force-Sensitive Manipulation” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Houzhu Ding, Xinda Qi, Luqi Wang, Ming Yang
CW-WAM predicts robot motion and contact wrench at the same future moments, then uses predicted wrench as a reference for admittance control to correct motion during execution. In 20 tight-clearance square-peg trials, it inserted the peg in 14 (70%), versus 5 (25%) for WAM without wrench input. These results cover four tasks within the demonstration distribution, with manually tuned gains and predefined compliant directions.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.