Understanding the behaviors of vehicles and other traffic participants at busy urban intersections is critical for urban planning, infrastructure development, and policymaking. Unfortunately, such understanding often comes after a huge investment for implementation and deployment. Many complex interactions occur infrequently and are difficult to capture through after-deployment monitoring. This project will develop a sensor-informed generative digital twin that integrates real-world data from the Riverside Innovation Corridor’s sensor network. By continuously integrating real-time sensory inputs, the platform can be used to create high-fidelity scenarios and simulate rare and challenging transportation dynamics. The digital twin will serve as a decision-support tool for policy evaluation, traffic efficiency strategies, and urban mobility planning. Its predictive capabilities will assist in designing infrastructure for autonomous vehicles, optimizing multi-modal travel demand, and enhancing energy efficiency. Through engagement with policymakers and stakeholders, the project will pave the foundation for the digital twin’s application in real-world decision-making. The proposed research will serve as a bridge, connecting data-driven insights with policy implementation towards sustainable transportation systems.