A Survey on AI- Based Smart PPE and Worker Safety Compliance System using Deep Learning
A Survey on AI- Based Smart PPE and Worker Safety Compliance System using Deep Learning
Dr.Rekha B Venkatapur
Department of Computer Science
and Engineering (HoD) K.S.Institute of Technology Bengaluru, India rekhabvenkatapur@ksit.edu.in
Aishwarya N
Department of Computer Science
and Engineering K.S.Institute of Technology Bengaluru, India aishwaryanarayana1@gmail. com
Akshaya B
Department of Computer Science
and Engineering K.S.Institute of Technology Bengaluru, India akshayab0502@gmail.com
Anusha V
Department of Computer Science
and Engineering K.S.Institute of Technology Bengaluru, India anusha28rev@gmail.com
Anvitha T A
Department of Computer Science and Engineering K.S.Institute of Technology Bengaluru, India anvithata21@gmail.com
Abstract— Ensuring worker safety in industrial environments such as construction sites, factories, and laboratories is a critical challenge worldwide. A large proportion of workplace accidents occur due to the absence or improper use of Personal Protective Equipment (PPE) such as helmets, gloves, safety vests, and masks. Traditional safety monitoring methods are manual, error- prone, and inefficient. This survey paper reviews recent approaches and deep learning-based techniques for automated PPE detection and worker safety compliance monitoring. The proposed system leverages YOLOv8- based object detection on video streams obtained from strategically placed surveillance cameras and, optionally, drone-mounted cameras. The system detects PPE compliance in real-time, monitors unsafe worker behaviour such as entry into restricted zones, and generates risk-based alerts. A dashboard displaying violation logs and compliance analytics assists safety supervisors in proactive decision-making. Experimental results from related work demonstrate detection accuracy exceeding 92% with low false alarm rates, indicating the viability of AI-based approaches for intelligent workplace safety enforcement. Keywords— PPE detection, YOLOv8, Deep Learning, Worker Safety, Object Detection, Computer Vision, Industrial Safety.