A GIS Approach to Dynamic Network Routing

Introductions
Recently, in transportation and GIS researches, the applications of Intelligent Transportation System (ITS) technologies have been seen as one of high potential directions to solve the problems of continuously degraded performance of transportation systems. By utilizing computers, communication and data processing technologies, the main goal of ITS is to improve the overall performance of transportation systems without building new highways.

ITS key programs cover a wide range of market areas that includes smart trrdlic control systems, emergency response systems, transit management systems, incident management systems, and traveler information systems. For these programs to succeed, the design models in ITS/GIS must simulate the real-world circumstances and must reflect reality. For the same reason, the embedded algorithm in each model of a real-time routing system has to function properly to MtIll user’s requests in terms of response time and information accuracy.

The trends of having reaJ-time in-vehicle navigation system on each car become much clear recently. For example, Mercedes Benz invests one billion dollars to design a Daimiler-Benz’s concept car in which the driver not only can send /receive e-mail and surf the world wide web, but can also receive real-time trrdlic information delivered from Traffic Operation Centers (TOC) (ITS World 1997). To offer more work productivity tools, convenience, safety and entertainment to millions of commuters who spend hours each day cruising the roads or stuck in traflic, at Comdex Fall 1997, IBM, Delco Electronics, NetsCape Communications and Sun Microsystems announced a technology that will utilize existing hardware and software technology -- including voice recognition, wireless communications, global positioning via satellite, head-up displays, Java technology, microprocessors, Web access and collaboration. and other Internet/intranet features -- to create a “Network Vehicle. ”

However, there are MTiculties that exist in applying GIS to this booming market. This paper will investigate the bottlenecks of applying current GIS technology in real-time vehicles routing, and thereafter, propose a conceptual model in GIS to represent and integrate data from various sources with dynamic traflic assignment theory from which accurate real-time or near real-time best route prediction result can be provided to driver for decision making.