JULIA: PPE USAGE MONITORING SYSTEM

Abstract

Effective and accurate monitoring of the use of Personal Protective Equipment (PPE) is fundamental for safety and compliance in industrial environments, where operations present a constant risk to the lives of employees and company assets. In this context, this work is characterized as applied, experimental, and exploratory research with the objective of describing the development and validation in an environment of JULIA, MVP of a SaaS (Software as a Service) system with a hybrid architecture, designed to perform automated monitoring of PPE use, offering a solution with a preventive focus regarding occupational safety. The system validation study was conducted in a simulated environment with approximately 200 training samples. The methodology included the development and application of the Student's t-test to verify if there was a statistically significant difference between the models and validation of the computer vision algorithm, using a predefined image database and the participation of the development team itself and third parties to assess the accuracy of the system in recognizing the use of PPE. The result was an algorithm trained to recognize the use of helmets, vests, and safety glasses with an accuracy of approximately 0.98. In short, it is understood that the personal protective equipment (PPE) monitoring system is capable of identifying with good precision the use of some essential equipment for high-risk operations.