Wireless Instant Fault Detection Through Finite-Element Trained Machine Learning

The rapid development of Industry 4.0 technologies has brought predictive maintenance into focus, particularly for small and medium-sized enterprises (SMEs) where cost and complexity are major barriers. In this paper, we present an innovative approach to vibration analysis, a key component for fault detection in mechanical systems and the creation of digital twins. Utilizing MatLab, we generated synthetic data points to simulate various vibration scenarios. These synthetic data points served as the training set for our machine learning model. The trained model was then integrated with a low-cost, Bluetooth-enabled accelerometer for real-time monitoring. Our system successfully identified fault conditions, specifically lump mass irregularities, through real-time sensor data. Our findings show promising capabilities for offering a cost-effective and straightforward solution for predictive maintenance. This research not only advances the field of vibration analysis but also opens doors for SMEs to embrace the benefits of digital twin technologies.