<![CDATA[PT XYZ is a company engaged in the rental of heavy equipment such as excavators, forklifts, bulldozers, and Wheel Loaders. The problem faced is unpreparedness in dealing with fluctuations in demand, so there is often a backlog or excess inventory. This research aims to improve the accuracy of demand forecasting and determine safety stocks to anticipate these uncertainties. The research was conducted using historical data on Wheel Loader rental requests from September to November 2024. The data was processed using the ARIMA (Autoregressive Integrated Moving Average) method through several stages, namely stationarity testing, identification of ACF and PACF, model estimation, parameter testing, white noise test, and selection of the best model. The resulting significant model was ARIMA (3,1,1), with a MAPE error value of 21% (79% accuracy), an increase of 9% compared to the previous method with an error of 30%. The results of the calculation of safety stock to deal with fluctuations in demand at various service levels show that the need for 2,688 units at the 90% level, increased to 3,444 units at the 95% level, and reached 3,900 units at the 97% level. This study shows that the ARIMA method is able to improve the accuracy of forecasting and provide a better basis for determining safety stock in managing fluctuations in heavy equipment rental demand.]]>
