Computer System with Neural Networks for the Management of the Computer Laboratory Sistema Informático con Redes Neuronales para la Administración del Laboratorio de Computación

Abstract

At Universidad Laica Eloy Alfaro de Manabí, El Carmen Extension, Software Engineering students rely heavily on computer labs for practical coursework, particularly programming. However, these facilities face a critical challenge: constant relocation of equipment and peripherals to accommodate diverse activities accelerates device wear-and-tear and shortens their lifespan. Although usage policies exist, staffing shortages and academic urgency lead to frequent non-compliance, resulting in uncontrolled inventory and rising maintenance costs. To address this challenge, we propose an intelligent system based on convolutional neural networks (CNN) that revolutionizes lab management. This system not only automates equipment allocation through genetic algorithms but also predicts technical failures using LSTM networks and tracks each device's location in real-time. Preliminary results demonstrate that this solution could reduce administrative time by 35%, optimize equipment utilization (currently at just 41.7%), and lower maintenance costs (which currently consume 27.6% of the annual budget). Beyond metrics, this project aims to enhance the educational experience: fewer disruptions from damaged equipment, improved security for technological resources, and a more efficient environment for students and faculty. As an adaptable model, it could be replicated at other universities facing similar challenges, particularly those with limited resources