Apple's HALO hits 91% success in real homes with zero real-world training
emmanuelvivier · x · 2026-11-11
Apple presents HALO (arXiv:2611.06908), a humanoid autonomy system trained on vision-language-motion data synthesized across a thousand procedurally generated homes — scaling with compute instead of human teleoperation hours. Skills include sitting, navigating to named objects, sidestepping through half-open doors, crouching under tables and pushing boxes, composed into long-horizon tasks. Without seeing a real home in training, HALO transfers zero-shot, achieving 91% success across 360 trials in six unmapped real environments. Code and videos are open.
More from Embodied
- Gemini Robotics humanoid serves snacks live on CoRL 2025 show floor — ki_ki_ki1 · 2026-11-11
- SuperMimic open-sourced at CoRL: mimics 129K motions with 1000x fewer GPU-hours than SOTA — kevin_zakka · 2026-11-11
- PolicyIK replaces numerical inverse kinematics with an RL policy, beating 4 classic IK controllers on a real robot — rsasaki0109 · 2026-11-11
- MIT's FLEX pipeline lets robots refine task representations from natural language corrections — andreea7b · 2026-11-11
- Liquid AI models are already running in Mercedes-Benz cars, phones next — agihouse_org · 2026-11-11
- AGI House hack day winners: on-device health, memory and research agent crews — agihouse_org · 2026-11-11