Identifying and Mitigating Congestion Onset (Phase 1)

This project created tools, empowered by big data, that can identify in real-time the onset of congestion and the occurrence of incidents, for both freeways and arterials. The target audience is highway system managers. The tools all track travel rates and flows, in real-time, towatch for trends that suggest the system’s operating status is changing; especially, transitioning from normal to congested operation either due to high traffic demand or an incident. They can use either vehicular data or roadside detector data, with corresponding adjustments to the processing procedure. This report describes what we have done to develop these tools; the ideas we tried; the ones that worked; to some degree, the ones that did not; the data we used; and the current status of tool refinement. Briefly, we used Bluetooth data from Sacramento, CA; probe data from Tampa, FL; and system detector data from Atlanta, GA. We tried about a half-dozen ideas; the three described are the best to date; and we are in the process of refining them. Our anticipation is that these tools will reduce the severity of the impacts from congestion and incidents because their occurrence will be detected sooner, especially for congestion, and more reliably. The report is available for download from the University of Florida at https://stride.ce.ufl.edu/wp-content/uploads/sites/153/2022/11/STRIDE-Final-Report-Project-J3-List-NCSU-compressed.pdf