Modeling The Best Route and Truck Capacity for Delivering Ice Cream Product: Vehicle Routing Problem Approach and Estimating Expected Interruption Cost Approach

摘要

Currently, the order processing of the studied company involves uploading the ice cream product into the truck and distributing it to various depots and stores after collected the orders. There were two problems faced by the studied company. The first problem is the distribution of products by drivers based on feelings and experience, so that the intended distance is not sequentially based on the closest and farthest. the second problem is the difference in truck capacity with delivery request for each depot, so it must be done twice at the time specified by each depot. The remaining product that cannot be delivered in regular hours, will be sent by overtime or will be postponed for the next day. Hence, based on this condition, the aims of this study is to find the optimal combination of the number of trucks related with their capacity and also to find a set of delivery regions and routes that satisfying some requirements or constraints and giving minimal total cost. as ice cream is a unique product that requires temperature -17 c, in order to minimize the transportation cost, this study prefer to find the optimal combination of the number of trucks rather than minimize the travel distance. Frequent opening of a truck door will cause sudden temperature increases, so the number of delivery locations for each truck cannot exceed three to prevent the products from melting. This study uses two approaches to resolve issues regarding delivery routes and truck capacity. This research uses a Vehicle Routing Problem (VRP) approach to consider the distribution scenario where a truck with specific capacity depart from the company, upload the ice cream products, and deliver them to a specific depot and store. This research also uses sensitivity analysis to see the impact of possible route failures on variable decisions, estimated costs and route reliability. Last, the research will use The Estimating Expected Interruption Cost (ECOST) method to estimate disruption cost that occ