This project develops a high-fidelity digital twin of the SMARTER Center’s Connected and Automated Vehicle (CAV) testbed at Morgan State University. The platform synchronizes infrastructure states, sensor observations, and traffic dynamics with a virtual environment in near real time, allowing safety and mobility interventions to be evaluated in a controlled, repeatable setting without exposing road users to risk.
By integrating live testbed data—including LiDAR, CCTV cameras, roadside units, and V2X messages—with simulation-based scenario testing, the project bridges a well-known gap in CAV safety validation. Physical testing is costly and slow, while purely virtual simulations often lack real-world calibration. This digital twin approach enables realistic scenario testing with validated sensor data, a capability that remains limited in existing CAV testbed configurations.
The platform demonstrates multi-modal capability through two application scenarios: pedestrian crossing conflict analysis at signalized intersections under varying speeds, visibility, and occlusion conditions, and transit signal priority evaluation using U.S. DOT bus trajectory data to assess potential operational impacts. The extensible architecture, with documented APIs, supports future applications such as emergency vehicle preemption, freight operations, and micromobility, and enables replication across diverse testbeds and agencies.
Universities Involved
Morgan State University
University of Maryland
Principal Investigators
Craig Scott (Lead PI)
Mansoureh Jeihani
Di Yang
Ehsan Mehryaar
Terry Yang
Expected Research Outcomes & Impacts
The project will deliver a functional digital twin system validated against physical testbed behavior, enabling virtual evaluation of safety-critical scenarios without road-user risk. Key deliverables include a 5-hour annotated dataset with DCAT-US metadata, a three-module software toolkit released via GitHub, and a scenario evaluation report with safety metrics and countermeasure assessments. The extensible platform design supports replication across other testbeds and future applications including adaptive traffic control, transit priority, and micromobility integration.
Transportation agencies and practitioners will benefit from a replication-ready model that enables data-driven safety evaluation and congestion mitigation. The project also trains two graduate students in CAV systems and digital twin technologies and delivers STEM outreach to 60 Baltimore City K–12 students through interactive demonstrations of digital twin and transportation safety applications.
Subject Areas
Connected and Automated Vehicles, Digital Twins, Traffic Safety, Sensor Fusion, Intelligent Transportation Systems