Adaptação do modelo de gestão de estoques em reposição periódica por meio do bootstrap para dados não paramétricos

Abstract

The traditional inventory management models are calculated based on some assumptions, among them, the assumption of normal data behavior. However, this premise is not always met, which can cause a disturbance when dimensioning stock levels. Thus, this study proposes an adaptation of the inventory management model in periodic replacement through the bootstrap to calculate the safety stock. To assess its applicability, demand data from 2013 to 2017 were used for blood components in a blood center in a city in the western region of Rio Grande do Norte. For that, the model performance adapted with the traditional one was compared from the perspective of four indicators, the models' performances were ranked and the Friedman test was performed to check for significant differences between the models, and later the discussion was carried out. the results and the applicability of the methods. The study shows that the traditional model obtained a better performance in meeting demand, however, the dimensioned stocks were often oversized, while the adapted model was less erroneous in terms of the difference between the dimensioned and the real, thus avoiding the oversizing of the stocks. In this way, the two models present their advantages and disadvantages, acting in a contrasting way with each other and presenting a trade off relationship, with the responsible manager evaluating the applicability of the best model to their reality.