Abstract
Increasing popularity of mobile route planning applications based on GPS technology provides opportunities for collecting traffic data in urban environments. One of the main challenges for travel time estimation and prediction in such a setting is how to aggregate data from vehicles that have followed different routes, and predict travel time for new routes of interest. One approach is to predict travel times for route segments, and sum those estimates to obtain a prediction for the whole route. We study how to obtain optimal predictions in this scenario. It appears that the optimal way to estimate travel times to minimize the expected mean absolute error is to combine the mean and the median times on individual segments, where the combination function depends on the number of segments in the route of interest. We present a methodology for obtaining such estimates, and demonstrate its performance on a case study using travel time data from a district of St. Petersburg collected over one year. The proposed methodology is meant for real-time prediction of expected travel times in an urban road network.