Shanghai-based Hesai Group, a lidar maker known for sensors used in autonomous vehicles, unveiled Kosmo, a mapping camera system that combines lidar and cameras to capture spatial data for training AI systems used in humanoid robots.

Hesai said Kosmo is a “3D World Data Engine” within a spatial intelligence platform for collecting and processing 3D world data for robotics and other “physical AI” applications.

CEO David Li said the robotics industry lacks enough high-quality spatial data and that robots still have limited real-world capabilities. He said Hesai had signed agreements with Chinese robotics firms to use Kosmo, but declined to give revenue forecasts.

The launch is part of Hesai’s broader robotics effort.

Sharpa Robotics, an independent startup founded by Hesai’s co-founders, raised about US$670 million.

Hesai’s actuators are used in Sharpa’s robotic hand, and Hesai said Kosmo’s mapping helped a robot using that hand achieve a 99% task success rate while serving ice cream at a Shanghai Dairy Queen outlet.

🔗 Source: Reuters

Recent Hesai developments