Sistema de monitoreo y respuesta para seguridad vehicular en tiempo real con tecnologías IoT e inteligencia artificial

Abstract

CONDITION FOR PUBLICATION OF PROJECT. current security situation in our country is one of the most critical problems, especially in transportation services. This project aims to implement an autonomous vehicle security system capable of detecting armed robbery or verbal intimidation, in order to generate and send real-time alerts to authorities for timely intervention. Project development included the use of the CAPDA methodology from data capture to the generation of final action alerts, based on the training of artificial intelligence models for correct detection and classification. The ESP32-CAM and ESP32 microcontrollers were integrated to use the built-in camera for image capture, and microphone and GPS modules were integrated to obtain complementary data. Furthermore, precision and accuracy tests of the models were favorable in terms of resources and inference times. The results showed the correct sending of processed data along with automated messages to the Telegram chat, keeping false positives to a minimum. Its autonomous design allows alerts to be generated and sent without manual intervention, improving the response capacity of authorities. Keywords: Vehicle safety, autonomous system, artificial intelligence, ESP32-CAM, ESP32, CAPDA.