Development of a Deep Learning-Based Framework for Real-Time Detection and Classification of Mechanical Faults in Rotating Machinery

Abstract

The early and accurate detection of mechanical faults in rotating machinery is critical for predictive maintenance and operational efficiency across industrial systems. This study proposes a novel deep learning-based framework that performs real-time detection and multi-class classification of mechanical faults in rotating machinery using vibration signal data. Our approach utilizes a combination of convolutional neural networks (CNNs) and long short-term memory (LSTM) networks to extract spatial and temporal features from time-series signals, offering enhanced performance compared to traditional signal processing methods. Evaluated on publicly available benchmark datasets, the model demonstrates superior fault classification accuracy and real-time response capability, thereby presenting a promising tool for integration into Industry 4.0 smart maintenance systems.