Real-Time Multi-Sensor-Based Data Acquisition System for Industrial Safety Monitoring

Abstract

Monitoring machinery is vital for detecting faults at an early stage before any hazardous incident occurs. A defect warning on the screen appears when vibration exceeds 6.5 meters per second squared due to an invisible bearing defect in machinery. Additionally, temperature sensors can detect overheating over time under these conditions. Otherwise, it remains normal and at a safe level based on the local visualization. This paper presents the development of a project and outlines the methods, including a design, tests, and equations used. Despite having multiple sensors affecting the performance speed, the system has satisfied real-time monitoring criteria, delivering the response of each output result within less than 35 milliseconds (0.035 s). This real-time data-acquisition-based electronic prototype closely aligns with industry standard principles and successfully recognizes both normal and faulty operations in industrial machinery. This 3-channel STM32-based project cut costs by 98% compared to a traditional and expensive 3-channel vibration monitoring device, helping corporations save costs and improve operator safety.