<![CDATA[General Background Modern agricultural machinery distribution centers face significant operational hurdles in maintaining continuous supply chains. Specific Background At the central warehouse of PT Corin Mulia Gemilang (MAXXI) in Sidoarjo, stockouts of fast-moving spare parts such as bimo oil seal rollers directly disrupt production lines and delay distribution to regional branches. Knowledge Gap Although classification systems categorize inventory movement, conventional replenishment models frequently fail to account for stochastic demand trends in agricultural machine components. Aims This study evaluates dynamic time-series modeling to forecast monthly demand for bimo oil seal roller components from August to December 2024 using historical records spanning January 2022 to July 2024. Results Application of the Box-Jenkins methodology demonstrated that data achieved variance and mean stationarity following third-order power transformations and first-order differencing, establishing ARIMA(2,1,0) as the optimal predictive structure with the lowest Mean Square Error of 73.672. Forecasted demand for the five-month horizon yields 934 units in August, 909 units in September, 887 units in October, 914 units in November, and 906 units in December 2024. Novelty The research demonstrates that time-series modeling effectively captures non-stationary, stochastic demand patterns specifically tailored to agricultural spare part warehouse logistics without relying on exogenous variables. Implications Implementing ARIMA-based forecasting enables warehouse managers to prevent inventory shortages, optimize safety stock levels, and preserve customer satisfaction across national distribution channels. Key Findings Highlights The ARIMA(2,1,0) model was identified as the optimal time-series model for predicting bimo oil seal roller requirements due to achieving the lowest Mean Square Error of 73.672. Demands for fast-moving agricultural components exhibit stochastic trend behaviors that require first-order differencing and power transformation to achieve statistical stationarity. Monthly demand projections for August through December 2024 remain relatively steady within the range of 887 to 934 units, indicating time-independent fluctuations. Keywords : Arima Model, Spare Parts, Inventory Forecasting, Warehouse Logistics, Fast Moving Items]]>
