A two-stage Metaheuristic approach for solving the Vehicle Routing Problem with Simultaneous Pickup/Delivery and Door-to-Door service

This paper presents a novel approach consisting of a two-stage optimization system to solve the problem of the last mile distribution of goods in the context of an Intelligent Transportation System (ITS). In particular, we deal with the Vehicle Routing Problem with Simultaneous Pickup and Delivery in conjunction with Door-to-Door operations (VRPSPD-D2D) to use vehicles travelling in the urban network to serve point-to-point shipping requests. The optimization problem is subject to some real-life constraints like vehicles' capacity, urban traffic flows, pickup and delivery sequences. The proposed optimization algorithm is based on two honey bee-inspired metaheuristics: Artificial Bee Colony (ABC) at the first stage; Bee Colony Optimization (BCO) at the second stage. This approach proposes an efficient way in achieving the main objective of this research work: the minimization of the global transportation cost, subject to the demands of the customers and operational constraints. To test our approach, new instances of the VRPSPD-D2D are proposed, based on the Sioux-Falls City network. Results of the proposed Bee System are compared with two Genetic Algorithm-based approaches.