Re:Figure 03, our 3rd generation humanoid robot
by eulgro on Thursday October 09, 2025 @05:53PM
(#45533505)
A more efficient way might be to train them in simulation. If you simulate a warehouse environment and use that to pre-train a million robots in parallel at 100x real time learning would go much faster. Then you can fine tune on reality for details missed by the simulation environment.
I’m sure they could pretty easily spin up a site with 200 of these processing packages of most sizes (they have a limited number of standardized package sizes) nonstop. Remove ones that it gets right 99.99% of the time and keep training on the more difficult ones, the move to individual items.
Caveat: I have no idea what I’m talking about.