Adaptive Forecasting And Genetic Algorithm For Order Planning Problem (Case Study: America Indochina Management Company)

Abstract

Forecasting for goods which are ordered from suppliers and executing order planning have always been in the top priorities of distribution businesses. Those problems regarding demand forecasting and order planning have become increasing dramatically because of many objective reasons: business’s warehouse capacity, short product life cycle, supplier’s risks, to name but a few. Following forecasting issues, order planning problem is also one of the big problems to be taken into consideration. This study determines problems regarding demand forecasting and economic lot-sizing with varying demand for a supply system consisting of suppliers and a distribution business. Adaptive forecasting techniques and an improving model for aggregate planning problem are established to solve their common issues that cause stockouts and project delays. The proposed model is applied for America Indochina Management Company Limited, a distribution company for FMCG (fast moving consumer goods), building materials, chemicals and office solution. This business faces demand from end customers. Historical data showed that demand changes over time, assuming it to be deterministic was too restrictive assumption. Hence, this study considers demand stochastics. Keywords: adaptive forecasting, order planning, aggregate planning, genetic algorithm, cplex software, stochastic demand.