Peramalan dan Pengendalian Persediaan Bahan Baku Kedelai yang Optimal dalam Produksi Tempe Menggunakan Metode EOQ (Economic Order Quantity)

Abstract

This study aims to determine an appropriate raw material forecasting method for UKM Marlianis, a household industry specializing in soybean processing to produce tempeh. Currently, UKM Marlianis adopts a conventional purchasing policy for soybeans without considering sales volumes, leading to overstocking or shortages of raw materials. The research employs a quantitative approach by comparing two forecasting methods: exponential smoothing and trend projection. The alpha (α) value used in the exponential smoothing method is 0.1. The findings reveal that the exponential smoothing method demonstrates lower forecasting errors than the trend projection method, with a Mean Absolute Deviation (MAD) of 374.38, Mean Squared Error (MSE) of 313,445.7, and Mean Absolute Percentage Error (MAPE) of 7.1%. Based on these results, inventory cost analysis was conducted using the Economic Order Quantity (EOQ) method. In conclusion, the exponential smoothing method with the specified parameters is more effective in forecasting soybean raw material needs, and inventory optimization using the EOQ method supports operational efficiency at UKM Marlianis.   Keywords: EOQ, Exponential Smoothing, Forecasting, Trend Projection