NdotLight has joined the Motif Technologies consortium for South Korea's independent AI foundation model initiative. The consortium plans to expand from a large-scale reasoning language model to VLM and VLA models while releasing model weights, code, and optimization libraries as commercially usable open source. Being entrusted with the 3D data layer of a national sovereign-AI initiative demonstrates NdotLight's recognition as a core technology supplier linking Korea's AI and robotics ecosystems.

NdotLight will lead the project's 3D data infrastructure for Physical AI training. Its role includes building an automated pipeline that generates manufacturing-grade 3D CAD from text and images and converts the results into SimReady formats for immediate use in simulation.

The company will also generate simulation-based synthetic data at scale for VLA models that integrate vision, language, and action. The goal is to reduce the cost of collecting real-world data while providing repeatable and scalable training environments.

NdotLight believes competitiveness in Physical AI will depend on how quickly and reliably high-quality SimReady data can be secured. The company will use the project to strengthen the data foundation connecting Korea's AI and robotics industries.