Apple's HALO: humanoid trained purely in simulation hits 91% success in real homes
heyshrutimishra · x · 2026-11-10
An Apple research team (authors include Vladlen Koltun and Matthias Müller) presents HALO: Perceptive Humanoid Motion from Simulation, attacking the humanoid autonomy data bottleneck:
- Data scales with compute, not human labor: teleoperation costs 1 human-hour per data-hour; HALO synthesizes vision-language-motion data across 1,000 procedurally generated homes
- Rich skill repertoire: sitting on chairs, navigating to named objects, sidestepping through half-open doors, crouching/crawling under tables, pushing boxes — orchestrated into long-horizon composite tasks
- Zero-shot sim-to-real: never saw a real home in training, yet achieves a 91% success rate over 360 trials across six unmapped real-world environments
- Hardware: Unitree G1 body; per the tweet, the "brain" is Alibaba's Qwen models running locally on a Mac mini strapped to the robot
- Code and videos released
More from Embodied
- At CoRL in Austin? Take a Tesla robotaxi ride, says robotics researcher — pduan · 2026-11-10
- Machenta demos instant robot learning at CoRL 2026, founded by Imperial College one-shot learning team — Ed__Johns · 2026-11-10
- NVIDIA Unveils GR00T 2 Open-World Action Model, Two Papers Shortlisted for CoRL Best Paper — danfei_xu · 2026-11-10
- Nomagic Team Headed to CoRL 2026 to Dig Into Physical AI Mastery — m_wulfmeier · 2026-11-10
- ROBIB Meetup #3 in Ghent to Debate Build-vs-Buy for Humanoid Robots — MarwaEldiwiny · 2026-11-10
- CoRL 2026 kicks off in Austin, Texas, running November 9-12 — shaohua0116 · 2026-11-10