Diseño del modelo de gestión de inventarios de baterías en la empresa LRV

Abstract

Here it is proposed a model of inventory management for automobile batteries at the LRV Company. Said model takes into account articles of type A with the objective of reducing the current level of sold out articles, as well as enhancing profitability, and customer service; there are neither prediction methods nor inventory politics, urgent requests are made permanently and delivery compliance is low. By using both, the ABC classification and the historical behavior of the demand to determine patterns, the articles of greater impact on sales were identified. Through simulation, and taking into account the error, the prediction method was chosen. The article-demand classification allowed the definition of the control system for each one of them, considering firstly, for the determinist demand, the EOQ and the time-varying demand models with periodic revision of the s-Q and s-S models. Secondly, for the probabilistic demand with continuous revision, the R-S and R-s-S models were defined. For an article with unpredictable demand, an adapted min-to-max system was formulated. This research showed that 22,8% of the articles provide 62% of the total sales. The best predictions for trend patterns are obtained through simple linear regression, and for perpetual patterns, simple exponential softening and mobile average with N=10 to 12. It was established that patterns are also obtained by formulating very high values for R that may be a bit unpractical.