Nvidia bets billions on physical AI, expanding Halos safety system to robots
Ars Technica AI · rss · 2026-10-08
Ars Technica reports Nvidia has bet billions on physical AI — robotics and self-driving — beyond its data center business, viewing safety as the next bottleneck. Halos launched in 2025 with hardware and software guardrails for autonomous vehicles; Halos for Robotics (announced June 2026) extends the safety architecture to warehouse AMRs, humanoid robots and even surgical robots. "The AI models are getting capable, the robot hardware is getting capable... safety is going to be the next bottleneck," said Amit Goel, Nvidia's head of robotics ecosystem.
More from Embodied
- Wandercraft calvin-40 carries 40kg, spotlighting robots' gap on heavy 'dull, dirty, dangerous' jobs — chris_j_paxton · 2026-10-08
- Eight Knot Shows Autonomous Boat Navigation with ROS at ROSCon JP — 4310sy · 2026-10-08
- POV: Cold Beer Delivered by Drone While Walking Around Shenzhen — Rahul_Chouhan · 2026-10-08
- MNV Humanoids exits stealth with ~$10M pre-seed; walking humanoid built in 5 months — lukas_m_ziegler · 2026-10-08
- Undergrad attends first IROS, gets hands-on with new humanoid robots — heatherknight · 2026-10-08
- Desktop robots MOSS and Vita star in an 80-minute robotics seminar at Thammasat University — StewartalsopIII · 2026-10-08