An optimization model and genetic-based matheuristic for parking slot rent optimization to carsharing

Abstract

Carsharing represents a major example of smart mobility service that allows a customer to rent a vehicle for a limited amount of time paying a per-minute fee. It may relieve people of the costly and non-sustainable burden of owning a car, especially when residing in a city. Though the spread of carsharing may bring significative benefits to (smart) cities, its penetration can be obstructed by non-up-to-date regulations, which can be still tied to a non-smart vision of mobility. In this study, we provide an overview of remarkable city regulations for carsharing, particularly highlighting the importance that parking policies can have in favouring the diffusion and use of carsharing services. Given such importance, we characterize the optimization problem of a local government that wants to analytically choose the best subset of parking slots to rent to carsharing companies, in order to improve urban mobility. To model and solve the problem we propose a new Binary Linear Programming problem and genetic-based matheuristic. Finally, we present results from computational tests referring to realistic data of the Italian city of Rome, showing that our optimization approach can return a fair territorial distribution of the parking slots, satisfying various families of constraints limiting the distribution.