SMART INDUSTRIAL PPE DETECTION SYSTEM USING YOLOv8
Lipika Chakraborty, Ashish Lal, Aayushi Bijawee
1. ABSTRACT
The Smart Industrial Personal Protective Equipment (PPE) Detection System using YOLOv8 is an advanced computer vision-based solution designed to enhance workplace safety in industrial environments. Industrial sectors such as construction, manufacturing, mining, and chemical plants are prone to hazardous conditions where safety compliance is critical. The failure to use PPE such as helmets, gloves, safety vests, and masks can lead to serious injuries or even fatalities. Therefore, automated monitoring systems are essential to ensure real-time compliance and reduce human dependency.
This project utilizes the YOLOv8 (You Only Look Once version 8) deep learning model, which is known for its high accuracy and real-time object detection capabilities. The system is trained on a dataset consisting of images of workers wearing and not wearing PPE. It detects multiple objects simultaneously and classifies whether workers are complying with safety standards. The system processes live video streams from surveillance cameras installed in industrial zones and identifies violations instantly.
The proposed system improves efficiency by providing real-time alerts when a worker is found without proper PPE. These alerts can be sent to supervisors through notifications or displayed on monitoring dashboards. The integration of artificial intelligence reduces manual inspection efforts and increases accuracy in detecting safety violations.
Additionally, the system can store data for further analysis, helping organizations identify patterns of non-compliance and improve training programs. The use of YOLOv8 enhances detection speed and precision, making it suitable for real-time industrial applications.
In conclusion, this project contributes to workplace safety by combining artificial intelligence and computer vision technologies to ensure continuous monitoring, reduce accidents, and promote a safety-first culture in industries.
