Real-Time IoT-Enabled Detection of Safety Gear Non-Compliance Using YoloV8 and OpenCV

This study presents a new method of enhancing workplace safety within industrial environments through the use of deep learning and Internet of Things (IoT) technologies. The research emphasizes timely recognition of deviations from the set Personal Protective Equipment (PPE) standards and compliance with tight safety guidelines that involve the compulsory wearing of Personal Protective Equipment such as arc suits. Our approach combines YOLOv8 object detection model with OpenCV and Arduino microcontrollers to detect workers who are not following safety guidelines with precision. A complete pipeline is adopted, including data capture, preprocessing, model training, and deployment. Data augmentation and hyperparameter tuning are used for enhanced model performance. The combination of IoT devices allows the system to issue real-time alerts, enabling the workers to be intervened upon right away and the potential risks minimized. Experimental findings prove the efficiency of our suggested method for enhancing industrial workplace safety [1], [2].