A “digital twin” traffic simulation tool for network-wide real-time traffic monitoring and management

Traffic monitoring is a cornerstone of traffic management and operations. To capture real-time traffic patterns and provide an instantaneous evaluation of traffic management strategies, this study proposes a “digital twin” traffic simulation framework using an open-source simulation package, namely “Simulation of Urban MObility” or SUMO to develop a real-time traffic monitoring and management system. Based on this framework, a simulation tool is developed to create a virtual representation of high-resolution network-wide traffic states that are updated from real-time traffic data (e.g., detector data). The tool is compiled into a Python package named DTSUMO for implementation. Compared to some existing traffic flow simulation tools, DTSUMO can simulate traffic patterns for short time windows depending on the detector data feed frequency, join vehicle trajectories estimated in consecutive time windows, and reflect real-time network conditions such as congestions and road closures. A case study is conducted using network and traffic data from Tuscaloosa, Alabama region. Results show that DTSUMO outperforms existing SUMO tools for traffic flow simulation, in terms of computational efficiency and simulation accuracy. The proposed tool takes advantage of existing traffic monitoring systems with loop detectors, and it can be extended with data from mobile devices and connected vehicles.