Role and Significance of Defect Detection Methods for Rotating Machines: An Explicit Literature Review

Defect detection in rotating machines reduces failure risks, minimizes operational downtime, and minimizes maintenance costs. This explicit literature review is presented from the research papers published between 2004 to 2024. It reports on vibration measurement and infrared thermography-based defect detection methods, with an emphasis on their theoretical foundations and practical applications. It is observed that substantial research is performed by various researchers on bearing defect detection, however, other defect types such as misalignment, gear faults, and unbalancing are given limited attention in the literature. It is imperative to address the aforementioned defects for sustainable maintenance of rotating machines. Vibration sensors are observed to be useful in detecting bearing defects by using characteristic time, frequency, and time–frequency detection techniques. Infrared thermography is an excellent and latest non-invasive tool for identifying overheating friction and lubricating issues by providing detailed temperature profiles leading to pre-defect identification in the rotating machine elements. It is observed that limited research is carried out on the combined use of vibration and thermography-based defect detection methods and their real-time applications. The successful utilization of combined vibration measurement and infrared thermography can yield faster, more accurate, and more precise defect detection, although managing the big data generated from multi-sensors may pose a significant challenge in the defect detection process. It is also evident that most of the published research in the defect detection field adopts standard datasets available online without validating their algorithm or program in real-time industrial applications. Recent evolutions in Artificial intelligence (AI) and Machine learning (ML) models have significantly improved the accuracy and processing speed of defect detection methods. AI and ML enable the automatic classification of defects and predictive maintenance by analyzing big datasets and evaluating the pattern of signals in comparison to the conventional defect detection methods. This literature review identifies the significance of various defect detection methods and vouches for the application of integrated and robust defect detection methods. It is also proposed to perform a real-time validation of defect detection methods by using AI or ML models to manage the big data generated from integrated approaches and improve the efficiency of data processing algorithms in the defect detection field.