Otimização do processo de compra: algoritmo e ferramenta

Abstract

This thesis report is part of the conclusion of the Integrated Master's Degree in Electrical and Computer Engineering of the Faculty of Engineering of the University of Porto and was made at Sonae SR, more specifically at Sportzone. This project's objective is to optimize the buying process of the upstream's team as well as making an implementation of a tool that helps in the buying process of products that come from suppliers. The objective of the upstream's team is to prevent stock failures as well as minimize stock days in the warehouse which is in Azambuja. This team controls stock needs of the warehouse and puts out orders to the suppliers according to those needs. Warehouse's stock needs come from sales from the stores as, as soon as it is generated a stock need due to sales in stores, naturally the warehouse sends the needed products reducing the stock days. The buyer's final objective is to maintain the stock days between certain levels considered acceptable by their bosses. In a first approach, the buying process was conceptualized, it's phases were defined as well it's constraints. It was noticed the stock failure is a main problem for retailing companies like Sportzone because it represents a missed purchase and/or a lost customer and, despite quantifying a missed purchase is easy, it is extremely difficult to estimate the cost of a lost customer in the long run as it might lead to missed future sales and lost in brand's trust. On the other hand, stock in excess implies a greater cost in stock maintenance being that is it necessary to take this aspect into account so as to we avoid unnecessary costs. This project rests on the optimization of the whole process of buying which has a lot of variables that condition its normal behavior. It is expected that the implementation of this automatic planning tool will reduce the time that each buyer must spend analyzing each article as well as prevent failures in article supervision.