Artificial intelligence software to detect microsleeps in drivers

• Autor: Morales Gonzales, Ruso Alexander
• Formato: artículo
• Fecha de Publicación: 2025
• Institución: Universidad de San Martín de Porres
• Repositorio: Revistas - Universidad de San Martín de Porres
• Lenguaje: español
• OAI Identifier: oai:revistas.usmp.edu.pe:article/2933
• Enlace del recurso: https://portalrevistas.aulavirtualusmp.pe/index.php/rc/article/view/2933
• Nivel de acceso: acceso abierto
• Materia: Artificial intelligence; Microsleep detection with AI; Facial recognition; Inteligencia artificial; Detección de microsueño con IA; Reconocimiento facial

This project was developed to address a significant social and economic issue: the detection of microsleeps in minibus drivers. Interprovincial trips, especially at night or during long journeys, increase the risk of fatigue and drowsiness, which affects the driver's attention. The consequences not only jeopardize the physical integrity of those involved but also lead to substantial material and economic losses. According to the World Health Organization, drowsiness-related accidents are one of the leading causes of mortality on roads worldwide (WHO, 2021). This project developed artificial intelligence (AI) software to monitor and analyze drivers' facial patterns in real time, identifying early signs of fatigue such as prolonged blinking, yawning, and head nodding. Ultimately, a mobile solution was achieved, capable of alerting the driver with auditory signals upon detecting a potential microsleep, allowing for early and effective interventions to prevent accidents. This technological advancement represents a significant contribution to road safety and can be extended to other public and private transportation contexts.