Aplicación de redes neuronales para la optimización de la cadena de suministro de la empresa Logex S.A.

Abstract

The purpose of this project is to analyze the application of artificial neural networks (ANNs) as an essential tool for optimizing the supply chain of LOGEX S.A., a logistics operator responsible for the distribution and storage of Cervecería Nacional’s cooling equipment. The main issue lies in the lack of accuracy when forecasting the demand for these cooling units in storage, which generates additional costs in spare parts, labor, and unnecessary mobilizations, and in some cases also results in contractual penalties. To address this challenge, a predictive model was developed in RStudio, based on a single-layer artificial neural network trained with historical data from 2023 to 2025, covering more than 10,000 requests for cooling equipment. The model was evaluated using statistical metrics such as RMSE, demonstrating superior performance compared to traditional forecasting methods by capturing non-linear patterns and significantly reducing estimation errors. The results show that implementing this predictive model improves resource planning, maintenance scheduling in storage, inventory optimization, and reduces operational costs. Consequently, this proposal enhances the efficiency of LOGEX S.A. and increases customer satisfaction for Cervecería Nacional.