Aplicação de métodos de previsão de demanda em uma empresa de serviços automotivos no Paraná

Abstract

The persuit of cost reduction and increased competitiveness has been a major challenge at the corporate level. Thus, the ability to forecast future data becomes a tool of great importance within a company. For the elaboration of this work, the company was divided between the sectors of air conditioners and radiators. From this, the three best-selling products in the last four years were selected, resulting in six data series. With this, a comparison was made between the methods of forecasting demand: Moving averages, Holt-Winters and ARMA. For three of the six products, being compressor, evaporator and additive, the model that presented the smallest error in the adjustment was the Moving averages, for the other products, which are condenser, radiator and plastic box, the model with the best fit was the ARMA.