A simulation-based inventory management with genetic algorithm for uncertain demand for third-party logistics provider

Abstract

This research aims to study and apply inventory management system for Third party logistics provider.
Currently, the company uses economic order quantity to control inventory. The analysis of historical demand
data shows that the demand is not deterministic. Hence, assumptions of using economic order quantity are
violated. In this research, the simulation-based technique is applied to solve for optimal order quantity and
reorder point. Since there are numerous items in the considered warehouse, ABC analysis is utilized to
select important items to analyze. Then simulation and genetic algorithm are applied to find the optimal
solution. Design of experiment with full factorial design is used to determine the best parameter setting of genetic algorithm. The performance measures are the average total inventory cost which composes of
average ordering cost, average inventory holding cost and average lost sale cost. The results show that the
average total cost for product code G2654, G2581, G0706, G2791 can be reduced by 73.43%, 49.86%,
28.50% and 13.38% respectively. For product code G2654, the average lost sale cost can be reduced by
85.30%. In summary, the solution from simulation and genetic algorithm provides better results than the one
from economic order quantity method.