Sensor Coverage and Location for Real-Time Traffic Prediction in Large-Scale Networks

The ability to observe flow patterns and performance characteristics of dynamic transportation systems remains an important challenge for transportation agencies, notwithstanding continuing advances in surveillance and communication technologies. As these technologies continue to become more reliable and cost-effective, demand for travel information is also growing, as are the potential and the ability to use sensor and probe information in sophisticated decision support systems for traffic systems management. This paper focuses on improving the efficiency of data collection in transportation networks by studying how sensor placement affects network observability. The objective of this study is to identify a set of sensor locations that optimize the coverage of origin–destination demand flows of the road network and maximize the information gains through observation data over the network, while minimizing the uncertainties of the estimated origin–destination demand matrix. The proposed sensor models consider problems where the numbers of sensors are limited and unlimited. The paper also provides several examples to illustrate the relative effectiveness of the proposed methodologies.