Just-in-time routing using real-time traffic data

We consider the problem of determining the departure time and point-to-point route for a pickup and delivery truck in a just-in-time environment. Assuming that the travel times along each arc in the network are random variables, we formulate a random horizon semi-Markov decision process model for the problem. Several solution techniques are discussed. We present several ITS policy and deployment issues that can be analyzed in the light of our model