Integration of Bluetooth, vehicle count data and trasport model results by means of Datamining techniques The application to the regional highway S.G.C. Fi-Pi-Li linking Florence to Leghorn and Pisa

The present article describes the research done in collaboration with the Department of Roads Planning and Project of Tuscany Region to study the traffic, and its main elements, for the main regional street corridor linking Florence with Pisa and Leghorn. The study integrates three different data sources, that’s to say, Bluetooth sensors data, traffic counts data and transport modelling results to study travel time, Origin/Destination and other information about this road. The methodology adopts an innovative Datamining algorithm called Predictive Association Rules to predict Origin/Destination matrix and integrates the transport model results to validate the algorithm and classification model elaborated. This is one of the first time that all these data sources are integrated and used jointly to study real traffic conditions.