Search

Transportation and logistics of valuable items (e.g., banknotes and coins, credit cards, securities, gold, jewelry, safes, and special pharmaceutical items like vaccines) are generally exposed to the risk of robbery and armored car heist. An inherent problem in designing supply chain networks for the transportation of valuables is to mitigate the risk while decrease the total cost by appropriate configuration of facility location and routing. In this research, the vehicle routing problem for the transportation of valuables is developed in several directions. From the modeling side, the risk mitigation approach that encompasses both the amount/number of valuables carried by a vehicle and the traveltime of a route is considered to minimize the risk of robbery and generate safe routes. The utilized risk function relaxes a pre-defined parameter of the risk threshold. Furthermore, the risk-minimizing problem is developed by incorporating various real-world characteristics and constraints into the model. At the expense of adding solution complexity, such formulation offers a more realistic model applicable to real situations.The model is extended from different angles, such as developing risk modeling by considering the vulnerability component and diversified arrival time using the multigraph network, and demand forecasting using an extensive evaluation of statistical and machine learning model. From the methodology side, multiple exact and metaheuristic methodologies are utilized and evaluated on several small to medium-sized instances and a case study. The augmented 휖-constraint 2 is used to solve the small instances of the problem. While solvable on small-sized instances, it poses computational challenges when applied to a large-scale rich problem. Therefore, five metaheuristics, namely, non-dominated sorting genetic algorithms (NSGAII and NSGAIII), strength of Pareto evolutionary algorithm 2 (SPEA2), indicator-based evolutionary algorithm (IBEA), and archived multi-objective simulated annealing (AMOSA) equipped with a novel solution representation and inconjunction with a modified Clarke-Wright saving algorithm are employed. The effectiveness of the proposed methods is evaluated and shown by examining various multi-objective performance measures. The case study is researched in more depth to obtain managerial insights. The results show that depending on the risk or cost efficiency of the solutions on a Pareto frontier, the risk of traversing longer routes or transporting larger amounts of cash can be determining in locating new bank vaults.