A new mathematical model for the service routing problem: A banking case study

In this study, problem includes pickup and delivery of a bank's office employees from specific stops by service vehicles, in the morning and evening. The service vehicle routing problem has been handled in many different ways in the literature. In terms of similarity to the problem, the Open Vehicle Routing Problem (OVRP) is discussed within the scope of the thesis. In OVRP, there is no return to the depot after the vehicle on a route has served the last customer. There are 171 node, including the workplace, in the problem that we consider the Asian and European sides of Istanbul together. The current workplace is located in an area with heavy traffic on the European side. The workplace is served by 54 vehicles with heterogeneous vehicle fleets from both Europe and Asia. In addition, the Gebze and Çayırova districts of Kocaeli, which are close to the Asian side of Istanbul, are also served by vehicles. OVRP with a heterogeneous vehicle fleet (HOVRP) is a NP-hard problem. For this reason, heuristic and metaheuristic methods have been used rather than exact methods in the literature to solve the problem. In this thesis, 2-index mathematical model proposed for a very difficult and real life problem. The 2-index model, the problem is expressed simpler and feasible solution can be produced even adding different constraints or increasing the number of nodes. The results obtained with the proposed model and the current solution are compared. Finally, the 2-index model was run for different operating times and the objective function values and the percentages of approaching the optimal solution were given according to the experimental design results.