Reproduce “BeyondContact: A Reinforcement Learning Framework for Airflow-Based Non-Contact 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
Peiyu Yang, Mohamad S. Moudallal, Antonio Langella, Giulio Turrisi, Victor Barasuol, Claudio Semini
BeyondContact adds experimentally identified airflow models to Isaac Lab and trains policies that combine pushing with blower forces for objects beyond direct reach. It models jet spread and applies surface-based forces and torques in simulation. In zero-shot hardware tests, the cube entered the target in 16/20 trials, while cuboid heading error averaged 2.6° across four tested targets. Results demonstrate these tasks only.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.