Nvidia Details South Korean Projects for Factories, Robots and Self-Driving

Nvidia is working with South Korean chipmakers, automakers and electronics firms on projects that show how it sees physical AI supporting industrial production, robotics and autonomous driving. The work spans factory monitoring, robot training and a planned self-driving system for South Korea.

At its AI Day Seoul event on Monday, Nov. 9, Nvidia said physical AI was entering a golden age and would be an important direction for the AI industry’s next phase. The company linked the shift to agentic AI, which it said is accelerating robot development, training and deployment.

Nvidia Vice President of Robotics and Edge AI Deepu Talla described robot development as a process involving model training, simulation and real-world use. Nvidia’s DGX, RTX and Jetson computing platforms, together with its Omniverse digital-twin technology, are designed to support that work from development through deployment.

Nvidia is working with Samsung Electronics and SK hynix on digital twins of semiconductor factories. The virtual models are intended to help monitor factory equipment and production processes. Hyundai Motor Group is using digital twins to train robots, with simulations capable of running hundreds of thousands of times faster than real time.

LG Electronics is using Nvidia’s GR00T platform to train its CLOiD robot to grasp and move objects. Nvidia also announced work with Kakao Mobility on a Level 4 autonomous-driving system for South Korea. The companies plan to integrate Nvidia’s Hyperion vehicle sensor and computing platform and Halos safety-validation technology into Kakao Mobility’s software, with initial development based on Kia PBV vehicles.