Willie A. Deese College of Business and Economics

Dr. Hyoshin Park: Anticipatory Traffic Sensor Location Problems with Connected Vehicle Technologies

Transportation Research Spotlight

Dr. Hyoshin Park: Anticipatory Traffic Sensor Location Problems with Connected Vehicle Technologies


N.C. A&T Assistant Professor Dr. Hyoshin (John) Park has received notification that a traffic sensor resulting from his research will be granted a U.S. patent. The research project, titled “Anticipatory Traffic Sensor Location Problems (SLPs) with Connected Vehicle Technologies” was conducted by Park and colleagues Ali Haghani, Song Gao, Michael A. Knodler, Siby Samuel and sponsored, in part, by CATM.

Dr. Park’s research expertise is in the area of decision making under uncertainty in transportation system analysis and network modeling. His current projects span several multidisciplinary areas in transportation engineering, including technologies in roadside sensors, emergency response, safety, focusing on operational learning in stochastic and dynamic systems and developing artificial intelligence (AI)-based decision tools. His AI algorithms have been applied to connected and autonomous vehicle operations sponsored by federal and state/local agencies including NSF, NASA, USDOT, VDOT, DOE, I-95 Corridor Coalition, Maryland Coordinated Highways Action Response Team (CHART) and NCDOT.

An example of a subarterial network in Burlington, N.C. where different numbers of sensors were tested in a total of 19 candidate locations.This research sought to offer a solution, responsive to uncertain and changing demand, to urban traffic congestion via the dynamic relocation and control of traffic sensors using wireless technology and portable devices. The basic idea was to have onboard equipment in vehicles transmitting data to roadside sensors integrated with traffic signal controllers. In response, the timing of traffic lights could be adjusted real-time to significantly reduce delays.

The research utilized two dynamic algorithm models to find solutions. The model framework was based on portable sensors that can be repositioned within the same day and from day-to-day to new locations so that locations of sensors and time stages of traffic lights are maximized. To tackle this problem, budget constraints on the sensor costs and relocation costs as well as various demand profiles and penetration rates of an urban transportation network had to be taken into account.