Computer System with Artificial Intelligence for Vehicle Security at the El carmen Extension University (ULEAM) Sistema Informático con Inteligencia Artificial para la Seguridad de los Vehículos de la Uleam Extensión El Carmen

Abstract

This article documents the development and implementation of an artificial intelligence-based computer system for vehicle security at the Universidad Laica Eloy Alfaro de Manabí (ULEAM), El Carmen Extension. Faced with the problem of inefficient access control and the community's perception of insecurity, the main objective was to design and implement a robust and scalable technological solution to automate vehicle identification and improve campus management. A quantitative-descriptive research methodology was used, with the waterfall development model guiding the project to ensure quality and sequential execution. The system's core component is an automatic license plate recognition (ANPR) model based on a YOLOv8 architecture, which was trained and optimized with a local dataset of Ecuadorian license plates to achieve optimal performance. The technical results obtained in a controlled test environment demonstrated the system's effectiveness, with a detection accuracy of 98.2% and a recognition accuracy of 95.1%. These values surpass the performance of similar systems in the literature. Furthermore, a survey of 150 members of the university community confirmed the need for the solution, showing that 73.3% perceive current security as "Regular" or "Insecure," while 96.6% believe an AI system would improve it. The efficiency analysis also yielded significant results, indicating that vehicle access processing time is reduced from an average of 15 seconds to just 3 seconds with the new system