Fasikaw Kibrete

Rotating machinery plays a vital role in modern industry and daily life, with widespread applications across sectors such as aerospace, manufacturing, and transportation. However, prolonged operation under harsh and variable conditions increases the likelihood of mechanical failures, which can lead to system‐wide disruptions, costly downtime, and serious safety risks. In recent years, artificial intelligence‐based methods, particularly hybrid deep learning approaches, have gained significant attention for their ability to provide timely and accurate fault diagnosis in rotating machinery. Among these, the integration of generative adversarial networks (GANs) and convolutional neural networks (CNNs) has shown promising potential. Despite growing interest, to the best of the authors' knowledge, no comprehensive review has yet been conducted focusing specifically on hybrid fault diagnosis methods that combine GANs and CNNs. This paper aims to fill that gap by providing a thorough review of current research on hybrid GAN‐CNN‐based fault diagnosis methods for rotating machinery. The review systematically synthesizes 217 peer‐reviewed publications, international standards, and authoritative technical resources identified using predefined inclusion and exclusion criteria. The literature search covered two major databases, Scopus and Web of Science, and spanned the period from 2010 to 2026. First, the fundamental principles and architectures of GANs and CNNs, along with their variants, are presented. Next, existing hybrid GAN‐CNN methods tailored for fault diagnosis are systematically reviewed, including their improvements and applications in the field. The paper then discusses the advantages and limitations of various GAN and CNN methods. Based on the presented review, existing challenges and future research directions are examined, followed by a conclusion.

Fault diagnosis in industrial equipment plays a crucial role in ensuring reliable system functionality and minimizing the costs associated with repair and maintenance. Traditionally, fault diagnosis has relied on human intelligence (HI), with skilled personnel applying knowledge and expertise based on reasoning and contextual understanding. Nevertheless, as modern industrial systems have grown more complex, operated at higher speeds, and generated massive amounts of data, relying solely on HI has become increasingly challenging and less scalable. Consequently, modern fault diagnosis has turned toward artificial intelligence (AI). This paper presents a comparative study of HI and AI in industrial fault diagnosis, based on a literature-driven analysis. The results confirm that while AI-based fault diagnosis systems perform well in processing large datasets and achieve improved diagnostic accuracy, these practices also face limitations related to data dependency, explainability, and deployment cost. By contrast, human intelligence remains indispensable in handling uncertain, rare, or new fault conditions that require contextual judgment and flexibility. The review further indicates that augmented intelligence (AuI) provides a collaborative framework that combines the complementary strengths of HI and AI for industrial fault diagnosis. Furthermore, emerging research directions, such as explainable and trustworthy AI, foundation models, large language models, physics-informed AI, digital twins, and human-centered AI, are identified as promising developments for next-generation intelligent diagnostic systems. The findings suggest that augmented intelligence is the most promising approach for advancing the performance and reliability of diagnostic systems in industrial machines.

Structural applications require materials that provide an optimal balance of strength, durability, and sustainability. This study develops functionally graded sisal/glass fiber-reinforced polymer composites and evaluates their tensile, compressive, and flexural properties, as well as their water absorption behavior. First, an experimental approach was adopted in which sisal fibers were treated with a 6% sodium hydroxide (NaOH) solution to improve fiber–matrix adhesion. The composites were fabricated using the hand lay-up method with varying sisal fiber ratios. Mechanical testing showed that the composites achieved a tensile strength of 146 MPa for the rectangular specimens and 199.5 MPa for the composite bars. The highest compressive strength (120 MPa) and flexural strength (326 MPa) were observed at 20% sisal fiber content. In addition, finite element analysis (FEA) results showed maximum tensile and compressive strengths of 181.24 MPa and 147.57 MPa, respectively, and were in reasonable agreement with the experimental results. Moreover, water absorption studies showed a direct correlation between fiber content and moisture uptake, with values of 1.40% (10% sisal), 2.55% (20%), and 3.85% (30%). The results indicate that functionally graded sisal/glass fiber-reinforced composites exhibit superior mechanical properties compared with conventional composites. The graded structure enhances design flexibility, optimizing material performance for structural applications. These findings highlight the potential of functionally graded composites as a sustainable alternative that addresses the limitations of traditional materials while promoting more efficient and environmentally friendly engineering solutions.

Natural fiber‐reinforced polymer composites are increasingly recognized as sustainable alternatives to synthetic composites due to their low density, renewability, and reduced environmental impact. In this study, alkali‐treated hybrid sisal/false banana fibers were used to reinforce a polyester resin matrix. The fibers were chemically modified via 5 wt.% NaOH alkali treatment to enhance fiber–matrix interfacial bonding. Composite laminates were fabricated using compression molding, and their tensile, flexural, and impact behaviors were experimentally evaluated for the designed hybrid configurations according to ASTM standards. Water absorption tests were conducted to evaluate the environmental durability, and finite element analysis was performed to simulate the tensile and flexural responses and validate the experimental results. The alkali‐treated hybrid composites exhibited significantly improved interfacial adhesion, resulting in tensile strengths of 13.67 ± 5.69 MPa (HYB‐AT‐1), 47.00 ± 12.17 MPa (HYB‐AT‐2), and 35.33 ± 3.21 MPa (HYB‐AT‐3); flexural strengths of 127.67 ± 5.25, 112.00 ± 4.00, and 117.67 ± 8.33 MPa; and impact strengths of 2.78 ± 0.05, 2.54 ± 0.53, and 5.31 ± 1.68 J for HYB‐AT‐1, HYB‐AT‐2, and HYB‐AT‐3, respectively. Among the hybrid configurations, HYB‐AT‐1 exhibited a flexural‐dominant performance profile, combining the highest flexural strength with moderate tensile and impact resistance and relatively low water absorption, indicating enhanced bending performance and environmental durability. HYB‐AT‐2 confirmed superior tensile strength, whereas HYB‐AT‐3 presented the highest impact resistance, reflecting the strong influence of fiber composition on property‐specific performance optimization. Good agreement between experimental and numerical results confirmed the reliability of the developed finite element model. These findings show that alkali‐treated hybrid sisal/false banana fiber composites can be tailored to achieve application‐specific performance, making them promising candidates for lightweight structural and semi‐structural applications.

Fault diagnosis of rotating machines has undergone significant advancements through the use of deep learning models. However, the effectiveness of these models is often compromised by noisy raw vibration data collected from industrial machines, which can negatively impact accuracy rates. To address this challenge, we present an improved fault diagnosis method that combines a multiple parallel hidden layer denoising autoencoder (MPHLDAE) with a one‐dimensional convolutional neural network (1DCNN). In this method, the MPHLDAE is used to effectively denoise the raw vibration signals, whereas the 1DCNN is responsible for diagnosing faults in the denoised data. Hyperparameter optimization was carried out through a grid search with k ‐fold cross‐validation. Experimental evaluations were conducted using rolling bearing and gearbox vibration datasets intentionally mixed with varying levels of Gaussian noise. The results validate the superior denoising capability of the MPHLDAE compared with existing denoising techniques. The proposed method consistently achieves fault diagnosis accuracies exceeding 95.30%, even under low signal‐to‐noise ratio conditions. These findings prove the superiority of the proposed MPHLDAE‐1DCNN method over other advanced methods for rotating machines.

Diagnosing faults in rotating machines is crucial for improving overall efficiency and productivity in modern manufacturing industries. However, limited fault samples and class imbalance issues impede the effective development and deployment of fault diagnosis models. Generative adversarial networks (GANs) have emerged as promising solutions to these challenges, and research in this field continues to grow rapidly. However, the latest GAN variants for vibration data generation in fault diagnosis have not been systematically reviewed. Moreover, the existing literature lacks a comprehensive evaluation of the quality metrics used to evaluate GAN‐generated vibration data and inadequately addresses the practical challenges faced in industrial settings. Therefore, this paper reviews the latest studies on the applications of GAN and its variants for generating vibration data in the fault diagnosis of rotating machines. The paper begins with an overview of the basic structure and principles of standard GAN and recent advancements, followed by a detailed analysis of their applications in generating high‐quality synthetic vibration data for fault diagnosis. The review paper then presents the metrics employed to evaluate the quality and diversity of GAN‐generated vibration data. Finally, this review paper presents practical considerations, discusses ongoing research challenges, summarizes recent advancements, and proposes directions for future research.

Rotating machinery serves as a critical backbone for national economic growth and is extensively utilized as mechanical equipment across diverse industrial domains. However, failures in these machines can cause significant operational disruptions, financial losses, and safety risks. This has led to an increased focus on advancing intelligent fault diagnosis techniques, with industrial big data and artificial intelligence (AI) emerging as powerful tools. Recently, deep transfer learning (DTL) has gained significant attention as a promising approach for cross‐domain and cross‐machine diagnosis, particularly in cases with limited faulty data and complex conditions. Despite significant progress, a comprehensive review of recent advancements in DTL remains absent, and well‐defined future research directions for its development are lacking. To address this gap, this review paper provides an extensive analysis of the existing literature regarding the recent applications of DTL in the fault diagnosis of rotating machines. This review categorizes existing DTL methods for machine fault diagnosis into four key aspects: instance‐based, feature‐based, parameter‐based, and adversarial‐based transfer. The survey first introduces the basic concepts of DTL, then explores recent research applications in depth and provides valuable perspectives in the field. In addition, the review examines the strengths and weaknesses of different DTL types in the fault diagnosis of rotating machines. Building on recent research progress, this review outlines emerging challenges and potential future research directions in the field. Thus, the review article aims to serve as a valuable resource and inspire further exploration among researchers, policymakers, and industry experts in the domain.