Workplace Safety with Personal Protective Equipment Monitoring System using Computer Vision

Abstract

Workplace environments, particularly in manufacturing plants, are highly exposed to occupational hazards, emphasizing the need for effective and reliable safety monitoring systems. Traditional methods of enforcing Personal Protective Equipment (PPE) compliance rely heavily on manual supervision, which is often affected by human fatigue, inconsistent judgment, delayed response, and limited monitoring coverage. These limitations reduce the efficiency of workplace safety enforcement and increase the possibility of occupational accidents and injuries. To address these concerns, this study developed a Workplace Safety with Personal Protective Equipment (PPE) Monitoring System Using Computer Vision designed to provide automated and real-time PPE compliance monitoring in industrial environments. The developed system integrates computer vision and deep learning technologies to perform visual data curation, personnel facial image registration and identity verification, PPE detection, compliance assessment, and real-time reporting. The system utilized cameras, microprocessor, facial recognition, object detection, centralized database storage, and a web-based dashboard for monitoring and analytics. Functionality testing results revealed that the visual data curation module achieved 93.33% accuracy, personnel identity verification achieved 96.67% stability, PPE detection obtained 92.11% precision with a recall score of 0.95 and an F1-score of 0.93, while the compliance assessment module achieved 90.00% effectiveness. Furthermore, the reporting module demonstrated 96.67% reliability in event logging and monitoring operations. The acceptability evaluation using the extended Technology Acceptance Model (TAM) obtained an overall mean rating of 4.68 interpreted as Highly Acceptable. The findings indicate that the developed system has strong potential to improve workplace safety management through automated PPE compliance monitoring, personnel verification, and real-time safety reporting.