An optimization model for vehicle routing of automated taxi trips with dynamic travel times

In this paper, we propose a method of automated vehicle operation in taxi systems that addresses the problem of associating trips to automated taxis (ATs) and assigning those vehicles to paths on an urban road network. This system is envisioned to provide a transport service within a city area with a seamless door-to-door connection for all passengers' origins and destinations. ATs can drive themselves on the roads with reduced direct human input, which allow taxis to satisfy the next trip or park themselves while waiting for a request if needed. We propose an integer programming model to define the routing of the vehicles according to a profit maximization function while depending on dynamic travel times which vary with the flow of the ATs. This will be especially important when the number of automated vehicles circulating on the roads is so high that will cause traffic congestion. The total profit involves the system revenue, vehicle fuel costs, vehicle depreciation costs, parking costs, penalties for unsatisfied trips and passengers' congestion delay. The model is applied to a small case study and the results allow assessing the impact of the ATs movements on traffic congestion and the profitability of the system. Even with a small case study, it is possible to conclude that having in consideration the effect of the vehicle flows on travel time leads to different results in terms of the system profit, the parking cost and the driving distance which points out the importance of this type of models.