Zhen Fan

About Me

I am an M.Sc. graduate from the National University of Singapore (NUS) and a Digitalization Engineer working on simulation-driven multimodal learning for embodied intelligence.

My research focuses on high-fidelity Digital Twins, world-model-based synthetic data generation, and Sim-to-Real perception systems for industrial robotics.

I work with NVIDIA Omniverse (Isaac Sim) and COSMOS to build scalable simulation pipelines, and I contributed to a Sim-to-Real action recognition framework validated at 90.31% accuracy in a real factory environment.

My broader research interest lies in combining simulation, multimodal models, and structured knowledge to enable explainable and safety-aware embodied AI, with long-term applications in Human–Robot Collaboration.

Research Areas: Digital Twin · World Models · Knowledge Graph · Sim2Real · Multimodal Learning · Embodied AI

View My Projects →