Condition monitoring in rotating machines: trends on application of Artificial Intelligence

Abstract

Due to the growing interest in enhancing industrial maintenance efficiency and reliability, condition monitoring has evolved through the integration of Artificial Intelligence techniques. In this context, advanced approaches have emerged for early fault detection in rotating machinery, emphasizing vibration analysis as the main information source. This article reviews current trends in applying Artificial Intelligence to condition monitoring, covering techniques such as machine learning and deep learning. Various methodologies are examined, including neural networks and hybrid models, designed to optimize anomaly identification in industrial equipment. Additionally, challenges related to system implementation are analyzed, including model interpretability, handling large data volumes, and integrating multiple information sources into the diagnostic process. Finally, future perspectives on these technologies are highlighted, stressing the importance of combining engineering expertise with data-driven models to enhance accuracy and reliability in industrial condition monitoring.