Freeway Incident Likelihood Prediction and Response Decision-Making

This research project consisted of two parts. The first part developed a set of real-time incident likelihood prediction models. The second part developed a freeway incident response decision-making methodology based on sequential hypothesis testing methods. The freeway incident likelihoods predicted by the real-time prediction models act as prior probabilities for the freeway incident response decision-making system. The products of this research project will be incorporated in the Advanced Traffic Management System that is being implemented on the Borman Expressway, a 16-mile segment of I-80 in northwest Indiana. The decision-making system can be used by traffic management personnel to assist in responding to various freeway incidents in a near optimal manner to minimize traffic delays and reduce the number of secondary incidents.