Crowd Sourced Data Organization and Analytics for Urban Traffic Estimation

Urban traffic estimation is a critical problem of various services in intelligent transportation systems (ITS) including optimal routing, improving transportation capacity. In order to accurately and timely estimate traffic condition (e.g. smooth, heavy or congested), real-time traffic data is crucial. This is a hard issue since urban traffic networks are large and complicated leading challenges for deploying fixed sensor systems at every location around the city, especially in developing countries where traffic infrastructures are not matured enough. This paper proposes a novel approach to crowd sourced data collection and analytics for traffic estimation. A prototype system has been deployed and evaluated. Real-field experimental results confirm the feasibility and efficiency of proposal approaches.