PROJECT NUMBER: PSR-26-09

FUNDING SOURCE:

TOTAL COST: 

PERFORMANCE PERIOD: JANUARY 2027 - DECEMBER 2027

Project description

As sensor technologies continue to advance, modern transportation infrastructure and connected vehicles are increasingly equipped with multimodal perception systems, including radar, LiDAR, cameras, and communication devices. The integration of these distributed sensors creates a cooperative sensing network that provides comprehensive and reliable situational awareness, enabling significant improvements in roadway safety and mobility. At complex intersections, fusing data from stationary infrastructure-based sensors with real-time sensor streams from passing connected vehicles expands sensing coverage, reduces blind spots, and provides critical system-level information. This collaborative integration of roadside and onboard sensors enables accurate tracking of vehicles, pedestrians, and cyclists, prediction of near-miss collisions, and rapid detection of localized anomalies. Collectively, advances in sensing technologies, artificial intelligence (AI), and cooperative perception create an unprecedented opportunity to significantly enhance the safety and operational efficiency of signalized intersections. 

The objective of this project is to develop a unified cooperative perception framework that aggregates and fuses multimodal data from both stationary infrastructure sensors and mobile sensing platforms to establish comprehensive, real-time situational awareness of complex intersections and demonstrate how this information can be used to improve safety and mobility. The proposed framework will be developed and evaluated using a complex intersection near the University of Southern California as a representative testbed. A high-fidelity three-dimensional (3D) digital twin of the intersection will be created that models the transportation infrastructure, traffic environment, all road users, surrounding buildings and obstructions, and the distributed sensing network, including both roadside and onboard vehicle sensors. The digital twin will serve as a real-time computational platform for tracking traffic participants, identifying potential near-miss events, generating proactive safety actions such as cooperative collision warnings or signal timing. The integrated sensing information will be used to optimize adaptive traffic signal control and cooperative traffic management strategies, with the objective of maximizing both safety and traffic efficiency under varying traffic conditions. The proposed research will demonstrate how the integration of cooperative sensing, 3D digital twin technology, and AI-driven decision-making can transform intersection management by providing a unified platform that simultaneously improves safety, mobility, and operational resilience. This framework will establish the foundation for future smart intersections capable of adapting in real time to changing traffic conditions while supporting increasingly connected and automated transportation systems and being applicable to a wide range of urban intersections.