🦾Standard Bots: $1B Valuation, Local Inference, Robot Models
TL;DR
Standard Bots' CEO Evan Beard and AI head Leif Jentoft explain how a $200M Series C company builds industrial robot models. Training runs in the cloud, inference stays on the factory floor, and models handle perception while classic code handles motion.
Standard Bots' CEO Evan Beard and AI head Leif Jentoft explain how a $200M Series C company builds industrial robot models. Training runs in the cloud, inference stays on the factory floor, and models handle perception while classic code handles motion.
Key Points
$200M Series C at a $1B valuation, led by General Catalyst and RoboStrategy
Customers named: NASA, Amazon and Lockheed Martin
Largest model is in the low billions of parameters; the perception backbone trained on over a billion images
In-situ fixes for edge cases can need only a few dozen examples
Simulation still struggles with liquids, suction and cutting flexible material
Why It Matters
Robotics AI is shipping as small, locally run perception models bolted to deterministic control. That pattern is more practical than end-to-end robot brains.
Quick Facts
Comments
Be the first to comment
Enjoyed this article?
Get it daily. 7am. Free. Reads in 5 minutes.