<![CDATA[This study evaluates the effectiveness of Distribution Requirements Planning (DRP) integrated with ARIMA time series forecasting to support delivery scheduling decisions and the determination of minimum inventory levels. As a representative case study, a 60-month sales series of Ultra-Pure Water was used to simulate fluctuating retail demand across the agent–depot network. The Augmented Dickey–Fuller test confirmed stationarity (p = 0.0142), allowing candidate ARIMA (p, 0, q) models to be evaluated using ACF/PACF and information criteria. The best model was ARIMA (1,0,1), which had the lowest Akaike Information Criterion and passed diagnostic tests (normal residuals, no autocorrelation, no heteroscedasticity), making it suitable for operational forecasting. Projection results indicated a stable demand pattern and yielded a safety stock threshold of 733.24 units/month (equivalent to 24.44 units/day) as a reference for inventory control. These findings demonstrate that the DRP–ARIMA integration can enhance supply reliability and distribution efficiency, particularly for subsidized goods such as 3 kg LPG, with practical implications for determining adaptive inventory levels, delivery routes and frequency, and upstream–downstream coordination. Theoretically, this study provides additional empirical evidence on the use of quantitative forecasting models to operationalize DRP in the energy sector, while also providing a foundation for replication in other critical commodities.]]>
Copyrights © 2025
