Skip to main content


Project Type
UTRC Advanced Technology Initiative
Project Dates
12/31/2008 - 12/31/2009
Principal Investigators
Project Status

Real-time traffic information, such as traffic flow, vehicle speed, vehicle classification, lane occupancy and so on, is important for traffic and vehicle control systems. Monitoring traffic flows on a transportation network is an essential topic in the field of traffic management and control. With the rapid advances of Intelligent Transportation Systems (ITS), a variety of sensors (e.g., inductive loop, ultrasonic, microwave, infrared detectors, and cameras) have been widely placed to fulfill this purpose.

Conventional sensors, for example, inductive loop, ultrasonic, microwave, and infrared detectors, have been put into use for several decades. Meanwhile video imaging vehicle detection systems are becoming an increasingly common means of detecting traffic at intersections and interchanges. This interest results from the recognition that video detection is often cheaper to install and maintain than inductive loop detectors at multilane intersections. Furthermore, an analysis of the data at Indiana showed that video detection was found to produce statistically significantly more false detections and missed detections than the loop detectors on most phases (Rhodes, et al., 2005). It is also accepted that video detection is more readily adaptable to changing conditions at the intersection (e.g., lane reassignment, temporary lane closure for work zone activities). The images obtained from the camera located beside and upon the road can be used for traffic surveillance, speed measurement, vehicle plate recognition, etc. Morita (1992) and Ozawa (1994) have shown that image sensors were widely applied in traffic and vehicle control. Furthermore, Yin et al. (2004) developed an algorithm using image sensors to measure real-time traffic flow parameters.

Many U.S. transportation agencies have adopted video vehicle detection technology as an effective alternative to inductive loops. For example, the New York City Department of Transportation (NYCDOT) places 98 cameras to watch over major arteries in the city and one Advanced Traffic Management System (ATMS) to control all the cameras and traffic signals to avoid congestion in the city. The collected real-time traffic information is broadcast to the public for their daily trip plan via internet. The Advanced Traveler Information System (ATIS) provides both streaming video and frequently updated still images from locations in the five boroughs.

Accurate and effective measurement and monitoring traffic are important to traffic management. In order to provide an accurate and reliable real-time estimate of the overall traffic condition and bottlenecks in a road network, significant number of sensors need to be placed over the whole network. However, the deployment of a large amount of sensors can be expensive and time consuming, given the constraint of the limited budget and time frame. Therefore, it is of great interest of studying an effective deployment strategy of sensors in order to install the minimal number of them while capturing the overall network traffic conditions relatively accurately. Furthermore, how to incorporate the information of existing sensors when developing an optimal strategy of installing new sensors remains a key issue for research and application purposes.

The goal of this research is finding an optimal sensor placement strategy to capture the overall performance of a transportation network effectively and efficiently. More specifically, this study aims at addressing the following questions:
1. What is the minimal number of sensors needed in order to reflect the overall traffic condition of the whole network to a certain confidence level?
2. Where each of sensors should be installed in order to achieve the above objective?
3. How to use the information from existing installed sensors to determine the optimal locations of new sensors?

In practice, this study provides an efficient and cost effective way of designing the sensor placement system and can be served as a valuable decision support tool to decision makers in public agencies in New York State.