A Genetic Algorithm for Solving the Inventory Routing Problem with Time Windows

Abstract

Logistics decisions of inventory allocation and transportation are strongly linked to each other since one directly affects the other, as is the case of how reducing the inventory in a customer facility may lead to a more intensive transportation process, because the smaller the quantities the more frequently the vehicles must visit the customer. For this reason, inventory and transportation decisions should not be considered separately but together, by using the inventory routing problem—IRP. This work presents a genetic algorithm to solve the IRP with time windows, which allows to simultaneously optimize inventory allocation and transport routes to supply a set of customers for a specific time horizon. This model allows to obtain a minimum total cost as a result of a better combination of the inventories at customers’ facilities and the transportation required to supply them. The proposed model and the algorithm developed for its solution allowed to obtain significant savings compared to the routing optimization to supply all customers in each period using the vehicle routing problem model with time windows, which allows to optimize customers’ inventory and minimize transport costs in each period. However, when this solution is compared with the total distribution cost throughout the time horizon, it generates higher costs than the solution generated by the IRP with time windows presented in this work.