Stanford's HomeBody lets a VLM-brained Unitree G1 clean an unseen kitchen
alex_verem · x · 2026-09-28
Stanford and Caltech's HomeBody system pairs a frontier VLM (GPT Astra) with an untrained Unitree G1 humanoid. In an unseen kitchen, the robot explores, builds spatial memory of the room, tidies up, and retrieves objects from underspecified requests — with no environment-specific training or new policy learning. The key bet: a System 2 VLM calling a composable skill library can replace the usual VLM→VLA→controller pipeline.
More from Embodied
- Sharpa's dexterous hand catches falling sticks at IROS 2026, beating humans in a live duel — YuXiang_IRVL · 2026-09-28
- AUAR ships robotic microfactories in containers to build homes up to six stories — lukas_m_ziegler · 2026-09-28
- GPT-6 Astra demo turns real-room video into interactive 3D worlds for robot training — 141_1337 · 2026-09-28
- Train robots for human kitchens, or redesign kitchens for robots? — vaibhavbetter · 2026-09-28
- PACTS at IROS 2026: Joint Action-Predicate Modeling Enables Zero-Shot Robot Skill Composition — joemeno · 2026-09-28
- Tactile-JEPA: topology-aware SSL cuts robot force error 6.3%, orientation error 20.8% — Elizaveta Kovtun · 2026-09-28