Reproduce “SwapBot: Tool-Centric Robot Manipulation via Interchangeable End-Effectors and Tactile Sensing” 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

Yunchu Zhang, Ray Song, Junhan Jiao, Liyiming Ke, Yuchen Yang, Ruihan Gao, Emma Romig, Siddhartha Srinivasa and 1 more

SwapBot equips a robot with interchangeable tactile-sensing household tools and learns tool-specific manipulation policies from demonstrations. A guided coupling enables tool changes, while the policies combine camera images, touch, and robot state to perform manually specified skills. In tofu preparation, full-task success was 60% over 20 rollouts, versus 35% for RGB-only; failures remained, and subgoals were manually specified.

Summary by alphaXiv.

Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.