Identifying and Mitigating Congestion Onset
George List, North Carolina State University
Billy Williams, North Carolina State University
Michael Hunter, Georgia Institute of Technology
Mohammed Hadi, Florida International University
Angshuman Guin, Georgia Institute of Technology
This project aims to create products that help transport agencies manage and mitigate the impacts of congestion. One product will be an algorithm that helps agencies detect recurring, incident, special event, or other related congestion sooner. System managers will be able to spot problems sooner and take preemptive mitigating actions. System users will experience less delay; agencies will use their financial resources more effectively; and society will be better off. The algorithm will capitalize on “big data” from various sources to identify the onset of incident, recurring, or other congestion. It will be trained to spot leading indicators of these events. The second product will be an algorithm (different from the first) that helps agencies engage in data-driven system performance management. This algorithm will use condition-based policy travel rates to spot network links and nodes that have the greatest need for policy-based mitigation. In both efforts, we will produce analysis tools that practitioners can use (stand-alone, prototype software intended to be integrated into existing system management platforms); as well as guide books for the use of the algorithms; and a project report. In the first year we will develop the analysis procedures (e.g., congestion “alarms”); and in the second year, we will fine tune these algorithms and recommend real time strategies that can use these tools to address impending or spreading congestion. The final products will be: 1) the algorithms, including code on GitHub, 2) guides for use of the algorithms, and 3) a final report.