Forecasting Demand of Moving Average and Linier Regression Methods in Predicting the Production of K93 Disc Brake Product (Case Study of PT United Steel Center Indonesia)

Abstract

PT United Steel Center Indonesia is a manufacturing company engaged in the steel plate industry. The
company produces various kinds of steel plate products and has various plants, one of which is the Disc
Brake K93 product. The Disc Brake K93 product is a part request from the largest customer which often
experiences significant changes, resulting in fluctuations and causing inventory to accumulate in the
warehouse. Inventories that accumulate can cause losses for the company, due to improper waste according
to planning, even less inventory can cause losses for the company because it can experience delays in
delivery to customers. Forecasting is one method in planning to predict and help draw conclusions.
Moving Average and Linier Regression methods are used in this study, and an analysis of the level of
accuracy is carried out using the concistency of MAD, MSE, and MAPE as well as forecasting validation
with Tracking Signal. Based on the results of processing and analysis, it is found that the linear regression
method is the most accurate method in predicting Disc Brake K93 products for the 2022 period with an
increase in production each month, the MAD value is 15242.23, MSE is 371649700, MAPE is 0.201, and
the Tracking Signal is 0.