Dr. Lili Du, associate professor in the Department of Civil and Coastal Engineering, has been working on several research projects in her years with the UFTI; some of her work has been supported by the NSF CAREER she received in 2016. The NSF CAREER award is the Faculty Early Career Development Program, which supports early-career faculty who have the potential to serve as academic role models in research and education and lead advances in the mission of their department or organization.
Dr. Du’s CAREER funded research looks at Coordinated In-Vehicle Routing Built Upon Online Learning and Distributed Optimization for Connected and Autonomous Vehicles.
The complex research Dr. Du is doing looks to solve real-life transportation issues. Her work seeks to tackle the increase in traffic congestion that many urban areas have been experiencing as a result of using routing guidance, like Google Maps or MapQuest. This increase in traffic congestion has negative effects on the function of urban transportation systems and a growing negative effect on the health of urban economies. Additionally, traffic congestion increases air pollution which negatively affects people and the environment as a whole.
Currently, two fairly accessible types of in-vehicle routing can potentially help drivers choose how to get from their starting point to their destination. One approach relies on independent routing where each driver can make their own routing decisions based on current traffic conditions. The other approach uses systemic routing which collects all drivers’ information and makes decisions for all drivers based on the best choices for the network as a whole.
Each of these approaches has its problems. Independent routing results in drivers making decisions based only on their own interest and results in oscillating traffic congestion for the network. Systemic routing may be best for the network, but it doesn’t always result in the best routes for an individual. In addition, it introduces a prohibitive computation load.
To address the deficits in these two approaches, Dr. Du proposes a new approach that coordinates the in-vehicle route choices among a group of vehicles as they travel. The approach relies on complex mathematical algorithms that can determine route decisions for all vehicles that opt in the service. This benefits the individual and the network.
Specifically, this coordinated routing system involves an information center that collects and aggregates traffic conditions along with starting points and destinations of individuals. Using complex mathematics the information center can estimate traffic conditions and send information back to individuals. Next, game theory is used to model the negotiation process that individual drivers use to make route decisions. Figure 1 below provides a visual representation of how this works. This coordinated routing system would happen continuously throughout a journey so that route choices can be re-coordinated and reevaluated as conditions fluctuate. Overall, this project develops novel in-vehicle routing approaches, which coordinate the route decisions of a group of vehicles online so that their collective decisions won’t worsen network traffic, while individuals can also be satisfied.