Sudo R1 Robot Entirely Trained in Simulation

It is no secret that researchers use a lot of real world data to train their models and robots. The Sudo R1 takes a different approach. It is trained entirely in simulation. As the researchers explain, there is no need for per-object tuning or calibration. As this video shows, the robot can pick objects it has not seen before, including glass, fabric, and irregular shape items.

A first look at what we’re building at Sudo AI. Introducing #sudo R1.

We train robots entirely in simulation, then deploy them directly to hardware – no data collection, per-object tuning, or calibration.

In this first release shared in our blog, #sudo R1 runs 60 minutes uncut… pic.twitter.com/vxDbnqTe8l

— Sudo AI (@sudo_robotics) April 20, 2026

[HT]