The two most important commodities of the 21st century are time and energy; traffic congestion wastes both. Several disciplines have studied the traffic congestion problem using mathematical models, simulation studies and field surveys. Recently, due to the sensor instrumentation of road networks and the availability of commodity sensors from which traffic information can be derived (e.g., CCTV cameras, GPS devices), a large volume of real-time traffic data at high spatiotemporal resolutions has become available. These data lay the ground for better understanding of the transportation network and for new applications.
The first objective of this project is to study the efficient acquisition and storage of traffic data. In particular, at the acquisition time, compact traffic patterns and outliers are constructed from raw data in real-time, supporting fast queries with approximate results and error bounds. The second objective is to utilize the traffic patterns to enable two critical query types that are building blocks of several important applications in time-dependent road networks: 1) shortest travel-time computation, and 2) k-nearest-neighbor search.
The broader impact of this project is to bring large amount of historical and real-time traffic data at the fidelities unseen before to the fingertip of transportation engineers, policy makers and end-users. This in turn enables them to alleviate one of the major problems of megacities, the main societal fabrics of the 21st century. Research results are incoporated into courses on Database Systems and Geospatial Information Management taught at University of Southern California and are also disseminated through papers published in conferences and journals. Details about the project including publications and open-source code are available at the project website: http://infolab.usc.edu/projects/TransDec/.
