DETERMINATION OF INVENTORY POLICY USING PROBABILISTIC APPROACH BASED ON QUANTITATIVE SALES FORECASTS IN PT KOMATSU INDONESIA

• Main Author: Estefina, Vania
• Format: Final Project
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/80123
• Institution: Institut Teknologi Bandung

Komatsu Limited is a Japanese multinational company that manufactures heavy equipment. Under the global brand of Komatsu Limited, Komatsu Indonesia produces complete machines such as hydraulic excavator, wheel loader, motor grader, dump truck, and bulldozer. These machines will be used for various sectors such as construction, mining, forestry, and agriculture. For the recent periods, Komatsu Indonesia’s two highest-selling products, PC130F-7 and D85ESS-2, have exceptionally high level of inventory, which leads to high costs. Currently, Komatsu Indonesia does not calculate their future sales and production number quantitatively. Therefore, in this research, estimation of the future sales and production number will be done with quantitative models by using exponential smoothing methods to determine future sales and probabilistic Q inventory model to determine production number as well inventory level. Based on the calculation of the forecast, it is shown that seasonal pattern plays a role in the data, hence triple exponential smoothing method will be used. This method also produces the least MAD among other exponential smoothing methods. The forecasted sales will later be used as an input to determine a new inventory policy. Using the probabilistic Q inventory model, the economic production quantity, replenishment point, safety stock, service level, and total cost (excludes production cost) will be the output of the calculation processes. Subsequently, production schedule will be shown to visualize the implementation of the new inventory policy. With the new inventory policy, there is a decrease of Rp198.267.808,00 that equals to 23,12% of the existing total cost (excludes production cost) in one-semester period.