Bring an agent or a robot policy. School trains it and lets it practise in measured worlds, then gives you the evidence that it got better in the real world, not just in the simulation. Pilot's policies are trained here.
An agent, or a policy for a robot, machine or vehicle.
nextLives and situations in worlds measured against reality, again and again.
nextEvidence that it improves in the real world, registered before it is tested.
next
Practice that transfers.
The worlds an agent learns in are scored against measurement, so what it learns there is worth learning.
Better, shown on the real thing.
Improvement is judged on real-world data the agent never trained on, with the bar set in advance.
A robot learns to cross a room through a Living Worlds crowd, thousands of practice runs at once in NVIDIA Isaac Lab on our own GPUs, by the same rules as the crowd it was tested against.
nowNVIDIA Newton and Warp for bodies, cloth, fluids and contact.
nowScenes rendered with NVIDIA Isaac Sim's RTX renderer, for what a robot's cameras would see.
nowLearned stand-ins for slow physics with NVIDIA PhysicsNeMo, scored on cases they never saw.
nowVideo from a prompt with NVIDIA Cosmos, always with its guardrail on.
nowNVIDIA Halos safety decisions, slow and stop, checked against Living Worlds crowds, with how late each one comes measured.
nowTell us what you want to run with School. We answer every request ourselves.