AI System for Detecting Safety Gear and Protecting Workers on Construction Sites

Reddypalle Rahul Reddy , Nagalakshmi Vallabhaneni , Siddareddy Reddy Srinivas , Siddareddy Reddy Venkatesh , Matta Sai Santosh

Artificial Intelligence, Computer Vision, Personal Protective Equipment (PPE), Safety Monitoring, Deep Learning, Object Detection, YOLO, Faster R-CNN, Worker Safety, Construction Site Safety, Real-time Detection, Machine Learning, Edge Computing, Cloud Deployment, Video Analytics, Automated Surveillance, Safety Compliance, Accident Prevention, Occupational Health and Safety (OHS), Workplace Monitoring, Human Detection, Smart Construction, Hazard Detection, Safety Management Systems, Predictive A

Construction sites are among the most hazardous working environments, with frequent accidents caused by the absence or improper use of Personal Protective Equipment (PPE) such as helmets, safety vests, goggles, gloves, and boots. Traditional safety monitoring is often carried out by supervisors through manual inspections, but this process is inefficient, prone to human error, and difficult to manage in large-scale projects where many workers operate simultaneously. In this context, the use of Artificial Intelligence (AI) offers a more effective solution for ensuring worker safety and compliance with safety regulations. The proposed system utilizes deep learning and computer vision techniques to automatically detect both workers and their safety gear in real time. High-resolution video streams from surveillance cameras are processed using state-of-the-art object detection models such as YOLO or Faster R-CNN. These models are trained on annotated datasets containing images of workers with and without PPE under various conditions, including different lighting, weather, and occlusion scenarios. Once a person is detected, the system verifies the presence of required safety gear by associating PPE detections with the corresponding worker. If non-compliance is identified, the system immediately triggers alerts through audio signals, on-site alarms, or notifications to supervisors. Beyond real-time detection, the system provides a centralized dashboard for site managers. This dashboard offers detailed compliance statistics, incident logs, and trend analysis to support decision-making and safety training initiatives. The solution can be deployed flexibly, either on cloud servers for centralized processing or on edge devices such as NVIDIA Jetson boards for low-latency, offline operation. The expected outcomes of this project include improved worker safety, reduction in accident rates, and enhanced operational efficiency on construction sites. By automating PPE compliance monitoring, the system minimizes reliance on manual inspections and ensures continuous, unbiased supervision. Furthermore, the recorded data and compliance reports can assist construction companies in meeting legal safety requirements and reducing insurance liabilities. In conclusion, this AI-based PPE detection system represents a step toward smarter and safer construction sites. Its integration of computer vision, real-time alerting, and compliance analytics provides a comprehensive framework for worker protection. With further enhancements such as helmet color recognition, fall detection, and hazardous zone monitoring, the system can evolve into a complete safety management platform, significantly contributing to the creation of accident-free workplaces in the construction industry.

Volume 13 Issue 11, November-2025

Pages : b645-b663