COMPARATIVE ANALYSIS OF MOVING AVERAGE AND DOUBLE EXPONENTIAL SMOOTHING METHODS FOR FORECASTING ASTM A252 GR 2 PIPE DEMAND AT PT XYZ

Abstract

Raw material inventory planning is a crucial aspect in the manufacturing industry to
ensure smooth production and cost efficiency. However, PT XYZ has not
implemented a forecasting method in its raw material planning system, so that
procurement decisions are still reactive to actual demand. This study aims to analyze
and compare forecasting methods using Double Exponential Smoothing (DES) and
Moving Average (MA) to determine the most accurate method in projecting raw
material needs for Non-API spec pipe products, type ASTM A252 GR 2 at KT 24 PT
XYZ. The data used is historical demand data, which is then analyzed using POM-QM
for Windows software. The results of the analysis show that the Moving Average
method with a two-month period (MA-2) has the smallest Mean Squared Error (MSE),
which is 182067, and a Mean Absolute Percentage Error (MAPE) value of 1.24%,
which indicates a higher level of accuracy than other methods. Thus, the MA-2 method
is recommended to be implemented in PT XYZ's raw material planning system to
improve production efficiency and reduce the risk of excess or shortage of stock. For
further research, it is recommended to develop a forecasting model by considering
external factors such as market trends and seasonality, and integrating machine
learning or hybrid forecasting methods to improve prediction accuracy. In addition,
the implementation of an Enterprise Resource Planning (ERP)-based system with a
forecasting module can also be a solution for long-term planning efficiency.