INTELLIGENT ADAPTIVE TRAFFIC FORECASTING SYSTEM USING DATA ASSIMILATION FOR USE IN TRAVELER INFORMATION SYSTEMS

In this paper we describe a novel application for data assimilation, a technique commonly used for atmospheric forecasting, in the area Traveler Information Systems. Our view is that the next generation of these systems will provide travel advise that employs not only the current traffic situation but also the fu- ture state of traffic. We show with actual traffic data that the impact of external influences (such as weather and large public events) are key to achieve accurate predictions. We propose an approach to model these influences in detail. Fur- thermore we describe how these detailed models can be combined with existing traffic models, using Data Assimilation. The results is an adaptive real-time sys- tem that can be used for intelligent Traveler Information Systems. We show our development and experimentation environment which offers solutions to the problems encountered in the development of the data assimilation system. Fi- nally we discuss the progress and the challenges ahead.