Reward AI launches OM-1: robot foundation model trained from 30 minutes of human demos
yifengzhu_ut · x · 2026-09-16
Reward AI came out of stealth with OM-1, a robot foundation model built on its Omnibody stack ('One Model, One Data Interface, Any Body'). Instead of teleoperation, humans wear the Omnibody Hand and perform tasks while the system records hand motion, tactile, proximity, vision, and position data. The same policy then controls different robot bodies, from industrial arms to humanoids. New tasks reportedly learn from under 30 minutes of human demonstration. Demos include four arms packing phones, folding laundry, and mixing cocktails.
More from Embodied
- UCLA's ChairNav teaches autonomous wheelchairs to learn each rider's preferences — zhoubolei · 2026-10-31
- Amazon Echo kills local voice storage: everything goes to the cloud from March 28 — alex_verem · 2026-10-31
- HP paid $116M for Humane, then bricked every $699 AI Pin 10 days later — alex_verem · 2026-10-31
- Moxie, the $800 AI robot friend for kids, went silent overnight after shutdown — alex_verem · 2026-10-31
- Subvocal's YC-backed glasses decode silent speech at 200+ wpm, built in 4 months — garrytan · 2026-10-31
- Nomagic warehouse robots go from bimanual to trimanual, pause to solve tricky picks — m_wulfmeier · 2026-10-31