Abstract:
lnventory data during January 2010 to December 2012 in a case study company
indicated that the material ordering quantity was much higher than material withdrawal
about 61 tons (about 4,823,000 baht). This inventory cost caused from forecasting
error. Previously, the company did an inventory forecast with the past experience. The
data of thirty six monthly materials uses during January 2010 to December 2012 are
used for forecasting in this factory case study. In this study, an ABC classification model
that analyzes a range of items and groups them into three categories (A, 8, and C)
is used to identify the highest cost of materials use that indicates the most important
items to minimize an inventory cost. Three materials (Copper, Silicon Carbide and
Carbon MA) are grouped in an A category accounting for 77.4% of the overall materials
uses in the company or about 8,053,000 bahts. A comparison of mean absolute
percentage error (MAPE) values from Moving Average, Single Exponential Smoothing
and Regression Methods is made. It reveals that MAPE value from Single Exponential
Smoothing method is the smallest and the most accurate one for this factory case
study. Economic Order Quantity (EOQ) and Reorder Point (ROP) are estimated.
Results suggest that the company can save over 1,110,733 THB over the past three
years if they applied the new EOQ and ROP policy.