Viswa Sri Rupa Anne, Georgia Institute of Technology
Yuqiang Ning, University of Florida
Lili Du, Ph.D., University of Florida
Srinivas Peeta, Ph.D, Georgia Institute of Technology
How can we mitigate peak congestion using smartphones and real-time incentives?
This project proposed to develop smartphone-based frameworks to develop/utilize real-time incentives (monetary, value-based, travel-related credits, information, etc.) to influence drivers’ en route routing decisions to manage network-level system performance in congested dynamic traffic networks. To achieve this objective, the project first investigated the role of demand management techniques in generating system-level benefits such as reduction in congestion or pollution, which explored two techniques, namely tangible incentives, and nudges. Both incentives and nudges were modeled in the context of network-level traffic congestion to be behavior consistent, real-time, and market-based. A reinforcement learning-based approach was employed to design and generate the incentives. A conceptual smartphone-based framework was illustrated to disseminate these techniques in the real world. Furthermore, this project exploited a novel information provision strategy to alleviate traffic congestion. It took advantage of the gaps information between individuals and the central planner (CP) to develop a correlated equilibrium routing mechanism (CeRM), which suggests priorities to individual vehicles’ route choices and drives their route choices to an equilibrium with a systematically optimal traffic condition, while still satisfying individuals’ selfish nature. A distributed Augmented Lagrangian algorithm (D-AL) was further developed to efficiently solve the CeRM to provide online real-time navigation services, taking advantage of the smartphones and/or onboard computation resources of individual vehicles.