Delta Introduces AI Digital Twin Platforms Using NVIDIA Omniverse

Delta Electronics announced in a press release at NVIDIA GTC that it has developed a series of AI-based digital twin applications using NVIDIA Omniverse to improve building automation and smart manufacturing solutions.

The company said its building automation platform integrates real-time data from lighting, HVAC systems, and environmental sensors with photorealistic simulations. This Omniverse-powered system can model physical conditions such as sunlight and heat distribution to optimize energy performance, achieving up to 20% energy savings potential. It also supports predictive maintenance and autonomous building functions.

In manufacturing, Delta’s DIATwin system uses NVIDIA Omniverse and OpenUSD-based models to create high-fidelity digital twins that simulate product design, robotics, and process data. Implemented in Delta’s Thailand facility, DIATwin enables faster production line design and validation, with NVIDIA PhysX providing accurate simulation of robotic operations. The use of NVIDIA Cosmos and Industrial Anomaly Generator further enhances automated optical inspection by generating synthetic defects for AI model training.

According to Delta, combining these digital twin systems with its Line Manager and Human Workstation platforms allows for decentralized manufacturing with centralized management, accelerating deployment and scaling toward autonomous factory operations.