Cálculo de estoque mínimo para itens industriais utilizando as distribuições de probabilidade normal, poisson e gama

Abstract

Efficient stock management of consumable items is very important for the proper functioning of any production chain. The considerable amount of items to manage, together with the changes in consumption patterns over time, become this activity quite complex. There are also logistical uncertainties, that is, regarding the availability of items. In this scenario, the determination of stock levels through merely empirical methods is impracticable, with a statistical approach usually being used. The present work uses Normal, Poisson and Gamma distributions, in order to calculate minimum stocks of items in a steel industry company. This project concerns the initial stage for the digital transformation of the supply chain of its industrial warehouse. Python programming language was used to organize data and perform statistical calculations, and PowerBI software to build the decision support system, to be used by warehouse analysts. Through the calculated minimum stocks, the system presented suggestions to increase the inventory of 30% of high turnover items, approximately. The system suggested the same value already used for minimum stock for 13% of items with very low turnover. The Gamma distribution was the most used among the low turnover items. The set under analysis is over 10 thousand items.