General Background: Monitoring personal protective equipment (PPE) usage is a critical component of occupational health and safety (OHS) in construction, yet manual inspection remains inconsistent and prone to error. Specific Background: Recent advances in computer vision, particularly YOLO-based object detection, have improved PPE detection accuracy in complex environments. Knowledge Gap: However, existing approaches primarily detect PPE presence without verifying its correct usage or associating it with individual workers, leading to inaccurate compliance interpretation. Aims: This study develops an automated PPE compliance verification system using YOLOv11l combined with spatial association logic to assess PPE completeness and anatomical correctness at the individual worker level. Results: The system was trained on 2,788 construction images and achieved high performance with mAP@50 of 0.979, precision of 0.976, recall of 0.954, and peak F1-score of 0.97, while demonstrating accurate classification across PPE categories including helmets, vests, and shoes. Novelty: The integration of zone-based spatial verification enables validation of PPE placement within anatomically defined regions, addressing the limitation of detection-only systems. Implications: This approach supports objective, continuous, and reliable safety auditing in construction environments, offering a scalable alternative to manual OHS monitoring. Highlights• Multi-class detection identifies workers and safety equipment with high accuracy• Region-based validation distinguishes proper gear usage from misplacement• System classifies compliance status through structured decision logic KeywordsAutomated PPE Verification; Construction Safety; Deep Learning; Spatial Association Logic; YOLOv11l