In many cities, traffic congestion is a serious problem due to limited roads and ever-increasing travel demands. Intelligent transportation systems (ITS) take advantages of the latest sensor, communication, and traffic control technologies to combat traffic congestion. These traffic surveillance technologies have been described as the eyes of ITS because they provide knowledge of existing network-wide traffic conditions. Effective traffic management and information systems rely heavily on a well-configured sensor network.

The objective of this dissertation is to develop sound mathematical models that systematically and optimally locate two types of traffic sensors in a transportation network. First, this dissertation addresses the traffic counting location (TCL) problem: to determine the number and locations of traffic counting stations for the purpose of origin-destination (O-D) trip table estimation. The proposed model emphasizes using available data including network topology, land use information, and O-D pair structure to determine an appropriate set of traffic counts that better capture the spatial distribution of O-D demands. A customized GIS tool combined with 3D visualization schemes is developed to graphically reveal the impacts of traffic counting locations on the quality of O-D trip table estimates. Second, this dissertation addresses the automatic vehicle identification (AVI) reader location problem: to locate AVI readers to capture a maximum number of trips and a maximum number of O-D pairs using a minimum number of AVI readers for the purpose of collecting travel time information. Variants of the AVI reader location model are also developed to consider time-of-day demand patterns. In addition, a methodology is developed for selecting the most preferred solution from the set of Pareto optimal solutions obtained from solving the multi-objective AVI reader location problem.

Two location problems in transportation are illustrated with case studies. The numerical results hold great promise (1) to achieve an equivalent level of accuracy, quality, and quantity of information while using fewer resources, or (2) to achieve a higher level of accuracy, quality, and quantity of information while using the same amount of resources.