The Ph.D. in Scientific Computing program is intended for students who will make extensive use of large-scale computation, computational methods, or algorithms for advanced computer architectures in their doctoral studies. This seminar series showcases the breadth of research covered by the program. 

Featured Speaker:

Xingmin Wang, PhD Candidate, Civil and Environmental Engineering and Scientific Computing

Xingmin Wang is currently a Ph.D. candidate in the Department of Civil and Environmental Engineering at the University of Michigan, Ann Arbor, advised by Professor Henry Liu. He obtained his bachelor’s degree in the school of vehicle and mobility from Tsinghua University, in 2018. His research interests include traffic state estimation and traffic network optimization with connected and automated vehicles.

Traffic signal optimization with connected vehicle trajectories

Traffic signal retiming is one of the most cost-effective methods for reducing congestion and energy consumption in urban areas based on the existing road infrastructure. However, high installation and maintenance costs of vehicle detectors have prevented the widespread implementation of adaptive traffic control systems (ATSC). Therefore, most intersections are still controlled by fixed-time traffic signals which are not updated regularly due to the lack of traffic monitoring capabilities. In the past few years, vehicle trajectory data has become increasingly available and offers many advantages over detectors and other infrastructure-based sensors for traffic monitoring; but using such data for automatic traffic signal diagnosis and optimization at scalable implementable levels is relatively unexplored. To fill this gap, this work proposes Optimizing Traffic Signals as a Service (OSaaS), an integrated traffic signal re-timing system that uses vehicle trajectories as the main input. OSaaS addresses many of the current challenges relating to signal retiming with trajectory data such as incomplete observation due to limited penetration rates. The system builds a queueing model that reconstructs the overall average traffic state, calibrated from performance measurements directly obtained from vehicle trajectories. The calibrated queueing model then predicts and evaluates network performance under different traffic signal parameters to provide diagnostics and direct traffic signal retiming guidance. In April 2022, a citywide field test of OSaaS was conducted in Birmingham, Michigan, with 34 signalized intersections. This resulted in decreases in both the delay and number of stops by up to 20% and 30%, respectively. OSaaS provides a more scalable, sustainable, resilient, and efficient solution to traffic signal retiming without requiring any additional infrastructure through the exclusive utilization of currently available trajectory data. As a result, it presents the possibility of upgrading all existing fixed-time traffic signals to dynamic systems with periodical parameter updates, something that is not currently possible without significant investments in infrastructure-based traffic flow sensors.

This event is part of MICDE’s seminar series featuring Ph.D. students in the Scientific Computing program. This series is open to all. University of Michigan faculty and students interested in computational and data sciences are encouraged to attend.

Questions? Email [email protected]