Abstract:
ABC (Pvt) Ltd is one of the largest entity in Sri Lanka which exports tires, tracks and wheels.
It is popular among its customers as a company delivering high quality products at competitive
prices. To achieve the production targets, the raw materials should be available on time. The
procurement department order raw materials from foreign and local suppliers and store them
according to the forecast done by using the System Applications Products (SAP) system which is
one of the largest enterprise resource planning softwares. The accuracy of the forecast done by this
method is 66 percent which leads to a high excess in the raw materials. In order to reduce this
excess, it is required to identify a forecasting method which gives high accuracy. The Box-Jenkins
and the Holt-Winters methodologies were suggested to find an accurate forecasting method. The
main raw material used for tire production is Carbon Black 2375. The monthly data was collected
from January ( 2012) to March (2019). The ARIMA (0, 1, 1) model is found as the best model in
Box-Jenkins methodology and it is compared with the existing forecast and the Holt -Winters
model and it shows the lowest mean percentage error. Therefore, ARIMA (0, 1, 1) was identified as
the best model with 79% of accuracy to forecast the inventory of Carbon Black 2375.
