ENERCOMP Seminar Series

ABSTRACT

The growing use of composite materials in industrial and energy systems necessitates advanced strategies for maintenance. Prognostics and Health Management (PHM) of composite structures in industrial applications using Artificial Intelligence (AI) provides a data-driven solution for detecting, classifying, and assessing damage. The proposed framework combines smart sensing, time-frequency signal processing, and AI-based models to extract critical health indicators. This work presents the state of the art in PHM across three domains: composite structures, power plants, and industrial robots. In composite structures, methodologies span statistical analysis, state-space modeling, and both shallow and deep learning approaches. Moreover, recent studies demonstrating the effectiveness of advanced hybrid and explainable AI techniques for accurate damage detection and diagnosis have also been included. For power plant applications, sensor selection using correlation analysis is included followed by the application of ensemble models and hybrid deep learning frameworks to monitor boiler tube leakage and turbine faults. For industrial robot PHM, the mechanical faults have been detected by machine learning using handcrafted features. Recently the process has been automated by using the autonomous feature extraction capability of the deep learning models. Moreover, to overcome limitations due to scarce fault data, physics-based simulations, and data augmentation techniques are employed. The results demonstrate that these industrial AI applications enhance diagnostic accuracy, automate decision-making, and reduce reliance on manual inspection. These approaches contribute to improving the safety, reliability, and operational efficiency of composite structures and energy systems through the implementation of AI-integrated PHM.