Definição e implementação de métodos de previsão de demanda na cadeia de suprimentos automotiva

Abstract

Companies in the automotive sector operate within complex supply chains, where inventory management represents both a strategic necessity and a financial challenge. Accurately forecasting the demand for components used in vehicle assembly is essential to balance parts availability with the reduction of operational and logistical costs. However, market variability, socioeconomic factors, and customer market strategies make this process uncertain. This study aims to analyze and apply demand forecasting methods in an automotive manufacturer, seeking to improve the accuracy of material planning. The research evaluates different statistical techniques, such as Moving Average, Exponential Smoothing, and the Box-Jenkins method, based on consumption data collected over four months of production. The results reveal that combining historical data with these methods provides a robust foundation for forecasting, with exponential moving average and exponential smoothing achieving the highest accuracy rates. However, optional items with high demand variability showed the most significant errors, indicating that economic factors and strategic choices directly affect forecast precision. Furthermore, external elements such as inflation, rising interest rates, and the devaluation of commodities have directly impacted consumer behavior and, consequently, demand forecasts. The study emphasizes that forecasting should not be treated as an isolated calculation but rather as part of an integrated strategy involving technical, commercial, and supply areas. It concludes that the adoption of robust predictive models, combined with improvements in information quality and forecasting methodologies, plays a decisive role in the company’s production and budgetary performance.