Reproduce “Heave-Ho: Decentralized Long-Horizon Multi-Humanoid Collaboration” 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

Ting-Hsuan Chen, Siheng Zhao, Han Jiang, Qi Pan, Louison Lu, Jiaheng Luo, Henghui Bao, Yiyu Chen and 2 more

Heave-Ho enables humanoid teams to move shared loads through coordinated local control and spoken skill commands. A leader selects each team skill, while robots control their own bodies and synchronize starts without network messages. In simulation across 12 HH-Bench tasks, it achieves 94.41% success versus 28.56% for Multi-HDMI. This measures low-level simulated task completion, not real-world success.

Summary by alphaXiv.

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