Robot sorting after 5h of training data, light box left OOD—diverse data is key
philfung · x · 2026-09-15
A developer shared new results in robot pick-and-sort training: 5 hours of data, the full box rather than just the top tray, and the light box deliberately excluded from training data as an out-of-distribution test. Not perfect and still makes occasional mistakes, but overall solid. The author's takeaway: diverse data is the key. Commenters asked whether work can continue after an error occurs.
More from Embodied
- Agility Robotics lost $138.1M in 2025 as Brett Adcock's doom prediction nears — Rewkang · 2026-10-31
- Promised the Year of Physical AI: Humanoid Robots Are Behind Schedule — Distinct-Question-16 · 2026-10-31
- PokeBot's RoboAPI brings the call-a-model pattern to robot arms, no task training needed — rohanpaul_ai · 2026-10-31
- Foundry Robotics raises $100M Series A to build factories that look like data centers — ishabytes · 2026-10-31
- NVIDIA's MotionBricks robotics motion-generation model trends on Hugging Face — nvidia · 2026-10-31
- Agility Robotics to raise $465M ahead of Nasdaq debut as AGLT on Nov 5, Foxconn leads with $201M — xmercury_one · 2026-10-31