PPE Compliance Monitoring using Lightweight YOLO Object Detection Models

The industrial and construction sectors require Personal Protective Equipment (PPE) compliance to protect workers from accidents at work, but the existing manual monitoring systems face challenges because they struggle to detect violations accurately. You Only Look Once (YOLO) based object detection models are used to create an automated system for real-time safety monitoring. The models are trained and evaluated on a publicly available PPE dataset containing annotated instances of safety equipment. The evaluation process involves testing lightweight YOLO variants such as YOLOv8n, YOLOv11n, and YOLOv12n under the same training conditions. The models are assessed using precision, recall, mean Average Precision (mAP) at IoU threshold 0.5 (mAP@50), and mAP averaged across multiple IoU thresholds from 0.5 to 0.95 (mAP@50–95). The results show that YOLOv12n detects objects with maximum accuracy because it performs better in various testing conditions than the other tested variants. This implies that the adapted lightweight version of YOLO can effectively perform real-time PPE compliance detection and can hence be used efficiently to monitor safety in construction sectors.