Forecasting Raw Material Demand for Battery Breaker Production Process

Abstract

This study aims to identify the most effective forecasting method for predicting raw material demand in the tin smelting industry, addressing the challenge of uncertainty in material arrival and inaccurate demand forecasting. Three methods, namely moving averages with n = 3 and n = 5, and exponential smoothing, were evaluated using historical data. Results indicate that exponential smoothing with α = 0.2 outperformed the other methods, yielding the smallest error rate with a Mean Absolute Percentage Error (MAPE) of 23%, Mean Absolute Deviation (MAD) of 411, and Mean Squared Error (MSE) of 293303. The implication of these findings underscores the importance of employing appropriate forecasting techniques to optimize inventory management and mitigate shortages in critical industries reliant on volatile raw material supplies. Highlights : Accurate demand forecasting is crucial for companies engaged in smelting to prevent shortages and inventory increases. Three methods were used to determine the most appropriate forecasting method for raw material demand based on historical data: moving average with n = 3 and n = 5, and exponential smoothing with α = 0.2. The Exponential Smoothing Method with α = 0.2 had the smallest error rate, with a MAPE value of 23%, MAD of 411, and MSE of 293303, and can be used to optimize demand forecasting for the next period. Keywords: demand forecasting, smelting, raw materials, historical data, moving average, exponential smoothing.