Proceedings of the 9^th International Conference on Operations and Supply Chain Management, Vietnam, 2019

Hashini Kodippili
Department of Transport & Logistics Management, University of Moratuwa

Nishal A. Samarasekera
Department of Transport & Logistics Management, University of Moratuwa

Effective route cycle design and assignment of vehicles in a fresh fruit and vegetable delivery network is of major importance since it contributes to a significant amount of transportation cost in the supply chain. Manual vehicle routing is a complex task due to geographical dispersion of customers, delivery time windows, vehicle capacity and delivery volume constraints. The objective of proposed method is to design an algorithm and develop a computer program which minimizes the total cost of travelled distance adhering to aforementioned constraints. When constructing the model, a background study has been conducted on the operations of the 3PL service provider and defined the objective function subjected to constraints. Inputs and outputs of the model have been defined to derive the cost formulas and then the vehicle routing and scheduling algorithm was designed. The model was proposed to a single depot distributing fruits and vegetables to multiple customer locations having deterministic demand. Routes are directed with known origin and destination where the route cycle is fixed once the journey begins. The study proposes an automated vehicle routing solution comprised of vehicle capacity utilization, departure time assessment and vehicle route assignment components. The total cost calculation formula is derived considering both direct costs and opportunity costs. It utilizes the cost per distance, cost of time window violations and cost of empty space of contracted fleet. The proposed solution was validated by comparing with the manual assignment of a 3PL service provider supplying fresh produce to supermarkets. The proposed method resulted in 26% total cost reduction and 40% reduction in total distance travelled.

[Download Full Paper] [Bibtex]

The persistent DOI of this article will be available soon .