Reproduce “Runtime Safety and Emergency Stops on a Full-Size Humanoid: Deployment Lessons and a Case for Graduated Force Authority” 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
Max Conway, Yutong Zhang, Allen Devaraj, William Xie, Alessandro Roncone, Nikolaus Correll
The paper examines a safety layer for balancing humanoids. The independent layer checks joint commands against limits and can trigger an emergency stop that cuts motor torque. In pushes, ALMI recovered positive center-of-mass margin in 14/15 trials, while MJPC required an operator stop each time. Evidence is limited to slow pushes on one H1-2, with falls caught by an overhead line.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.