Connected Traffic Intelligence

Connected vehicle (CV) data has transformed how agencies measure traffic signal performance, helping identify where delays occur. But knowing where congestion exists doesn’t always tell us where retiming will actually help. Many problem intersections are constrained by operational limitations, and today’s tools lack the ability to predict how different timing strategies will perform before implementation. As a result, CV data is often under-used for proactive signal optimization.

To address this gap, researchers at the University of Michigan developed a traffic modeling approach that works even when connected vehicle penetration is low. By using the trajectories of a small sample of vehicles, the model estimates traffic demand without traditional turning-movement counts. In an initial deployment in Birmingham, Michigan, the system reduced average delay by 20% and stops by 40%—without installing any new roadside detection. The system is now being prototyped through a county-wide deployment with the Road Commission for Oakland County, funded by a Stage 1 SMART Grant from the U.S. DOT.

Connected Traffic Intelligence (CTI) builds on this research to deliver a cloud-based signal optimization platform that continuously retimes signals using only existing CV data.

“This tool addresses a critical gap in our practice by providing unprecedented insight into our traffic signal control performance. It offers significant opportunities to enhance our system by reducing congestion and emissions while improving safety, even with limited resources. We are confident that this technology will drive the next generation of traffic signal control, with the potential for nationwide implementation.”