Predictive Maintenance in Industrial Equipment Using IoT Devices

Abstract

The rapid evolution of Industry 4.0 has necessitated innovative solutions for maintaining industrial equipment to enhance operational efficiency and minimize downtime. This thesis explores the development and implementation of a predictive maintenance system using Internet of Things (IoT) technologies, with a focus on real-time vibration monitoring and classification. The study employs the Nicla Sense ME sensor and Arduino MKR WAN 1310 microcontroller, integrated with the Arduino IoT Cloud, to monitor vibration data from a PC fan as a representative industrial component. The project demonstrated reliable performance and achieved significant results, demonstrating a functional IoT-based predictive maintenance system capable of identifying vibration anomalies in real-time. However, the implementation faced challenges, including sensor integration issues, environmental noise interference, communication delays, and limited computational resources. Despite these limitations, the findings underscore the potential of IoT technologies to transform maintenance practices by enabling predictive insights and optimizing resource utilization. This work highlights the feasibility of scalable, IoT-driven predictive maintenance systems and identifies avenues for future improvements in hardware compatibility, data analytics, and machine learning integration. By addressing these areas, predictive maintenance can redefine industrial equipment management, contributing to increased efficiency, reduced downtime, and the advancement of smart manufacturing practices in the era of Industry 4.0.