Building a real-time traffic and congestion surveillance system based on taxi GPS data

This paper is a contribution toward an operational use of large floating car data (FCD) in traffic management. The work focused on a practice-ready traffic surveillance system for Chongqing City of 8 million people. In this system, the real-time GPS FCD of all the 16,000 taxis in the city are analyzed to monitor the traffic performance in the city in different levels, including road segment level, corridor level, sub-area level, and city level. In this paper, the MOE selection, system outcomes and related recommendations on future system improvements are discussed. In this system, travel time ratio has been selected as the key MOE to evaluate traffic performance since it incorporates many factors that affect travel time and it is very intuitive and easy to understand for the public. It is demonstrated that using large-scale of FCD data to monitor traffic performance is a feasible and cost-effective solution to detect and better manage transportation.