Helm.ai signs $70 million in physical-AI model contracts

Helm.ai's roots are in autonomous driving, which the company says requires generalizing across different environments such as busy city streets or desert roads. That experience provided the foundation for expanding into other areas. Voroninski said the company has shown its technology can generalize to entirely different application areas, using the same perception stack for autonomous driving, an open pit mine, or industrial environments.

The company also learned from the automotive industry that real-time latency constraints and rigorous safety regulations are part of the package. According to Voroninski, being able to handle that industry makes Helm.ai well equipped for others. The approach separates understanding an environment from acting in it, which the company says helps with data efficiency and safety certification.

The commercial traction comes as robotics is, in Voroninski's words, almost where autonomous driving was 10 years ago. He said robotics is now entering its breakout moment and that Helm.ai is interested in working with robotics companies across industries and embodiments.