Maintenance Applications Research Articles

A Hybrid Deep Learning Model for Enhanced Structural Damage Detection: Integrating ResNet50, GoogLeNet, and Attention Mechanisms

Quick and accurate structural damage detection is essential for maintaining the safety and integrity of infrastructure, especially following natural disasters. Traditional methods of damage assessment, which rely on manual inspections, can be labor-intensive and subject to human error. This paper introduces a hybrid deep learning model that combines the capabilities of ResNet50 and GoogLeNet, further enhanced by a convolutional block attention module (CBAM), proposed to improve both the accuracy and performance in detecting structural damage. For training purposes, a diverse dataset of images depicting both structural damage cases and undamaged cases was used. To further enhance the robustness, data augmentation techniques were also employed. In this research, precision, recall, F1-score, and accuracy were employed to evaluate the effectiveness of the introduced hybrid deep learning model. Our findings indicate that the hybrid deep neural network introduced in this study significantly outperformed standalone architectures such as ResNet50 and GoogLeNet, making it a highly effective solution for applications in disaster response and infrastructure maintenance.

Research on damage identification of simply supported bridge based on effect size method for vehicle-bridge coupled vibration

Abstract This paper presents a novel bridge damage identification method employing Cohen's d as an effect size indicator, predicated on the detection of bridge damage through the coupled vibration between a vehicle and the bridge. By analyzing the dynamic response of the bridge as a vehicle passes over, this method effectively extracts the modal parameters of the bridge and facilitates the identification of bridge damage. Numerical models of the bridge under various damage conditions, including no damage, mid-span damage, and damage at the 1/4 and 3/4 span locations, have been constructed to substantiate the efficacy and precision of the effect size as a damage indicator. Furthermore, to address the challenge of accurately identifying damage at boundaries, which is often confounded by boundary effects, a boundary subdivision method has been defined for the detection of boundary damage. Through the implementation of a real-bridge test based on indirect measurement technology, the self-vibration frequency of the bridge was successfully extracted. This empirical data was then compared and analyzed against results from ANSYS numerical simulations, thereby validating the practicality and accuracy of the proposed method. In the final analysis, the influence of random traffic flow on the bridge damage identification results was examined. The findings indicate that the vibrations in simply supported beam bridges are intensified due to the impact of random traffic flow, which aids in enhancing the accuracy of damage identification. The introduction of this method provides a new quantitative tool for bridge health monitoring, enabling the rapid and accurate identification of bridge damage without interrupting traffic. It holds significant value for engineering applications in bridge maintenance and safety assessment.

Generative AI for Predictive Maintenance: Predicting Equipment Failures and Optimizing Maintenance Schedules Using AI

Predictive maintenance has emerged as a transformative approach to managing equipment health, reducing unplanned downtime, and extending asset lifespan. Leveraging advancements in generative artificial intelligence (AI), this paper explores the role of AI-driven predictive maintenance in predicting equipment failures and optimizing maintenance schedules. Traditional maintenance strategies, such as reactive and preventive approaches, often lead to inefficiencies, increased operational costs, and unexpected breakdowns. Predictive maintenance, powered by AI, offers a proactive alternative that not only anticipates failures but also enhances scheduling efficiency, maximizing equipment uptime and reducing maintenance costs. Generative AI models, including techniques such as Generative Adversarial Networks (GANs) and reinforcement learning, have shown immense promise in learning complex patterns from historical data and simulating potential equipment failure scenarios. These AI-driven models can analyze vast and diverse data sources—including sensor readings, maintenance logs, environmental conditions, and historical failures—to provide accurate, real-time insights into equipment health. This paper details the architecture and functioning of generative AI models in predictive maintenance, emphasizing their role in both anomaly detection and failure prediction. A systematic comparison of reactive, preventive, and predictive maintenance is provided, underscoring the unique benefits and challenges of predictive maintenance. We discuss the types of data essential for predictive maintenance and present sample data structures used in model training and deployment. Additionally, this paper demonstrates how generative AI models predict equipment failures by identifying anomalous behaviors before they escalate, enabling preemptive actions. A failure probability model is presented to illustrate how failure risks evolve over time, alongside tables showcasing the critical data points in predictive maintenance. The paper also explores the optimization of maintenance schedules using generative AI, where models simulate and compare different maintenance timing strategies, ultimately minimizing downtime and maximizing productivity. However, we also acknowledge the current limitations of generative AI in this domain, including data privacy concerns, computational intensity, and the challenges of model interpretability for practical implementation. Looking forward, we examine future trends such as the integration of Internet of Things (IoT) devices and the emergence of more sophisticated AI models that will likely enhance predictive maintenance applications. This paper concludes by highlighting the transformative potential of generative AI for predictive maintenance, offering insights for industries seeking to innovate their maintenance practices and achieve superior operational resilience.

Challenges of the Biopharmaceutical Industry in the Application of Prescriptive Maintenance in the Industry 4.0 Context: A Comprehensive Literature Review

The biopharmaceutical industry has specificities related to the optimization of its processes, the effectiveness of the maintenance of the productive park in the face of regulatory requirements. and current concepts of modern industry. Current research on the subject points to investments in the health area using the current tools and concepts of Industry 4.0 (I4.0) with the objective of a more assertive production, reduction of maintenance costs, reduction of operating risks, and minimization of equipment idle time. In this context, this study aims to characterize the current knowledge about the challenges of the biopharmaceutical industry in the application of prescriptive maintenance, which derives from predictive maintenance, in the context of I4.0. To achieve this, a systematic review of the literature was carried out in the scientific knowledge bases IEEE Xplore, Scopus, Web of Science, Science Direct, and Google Scholar, considering works such as Reviews, Article Research, and Conference Abstracts published between 2018 and 2023. The results obtained revealed that prescriptive maintenance offers opportunities for improvement in the production process, such as cost reduction and greater proximity to all actors in the areas of production, maintenance, quality, and management. The limitations presented in the literature include a reduced number of models, the lack of a clearer understanding of its construction, lack of applications directly linked to the biopharmaceutical industry, and lack of measurement of costs and implementation time of these models. There are significant advances in this area including the implementation of more elaborate algorithms used in artificial intelligence neural networks, the advancement of the use of decision support systems as well as the collection of data in a more structured and intelligent way. It is concluded that for the adoption of prescriptive maintenance in the pharmaceutical industry, issues such as the definition of data entry and analysis methods, interoperability between “shop floor” and corporate systems, as well as the integration of technologies existing in the world, must be considered for I4.0.

Daily Engine Performance Trending using Common Flight Regime Identification

Abstract Accurate flight regime identification is critical for enhancing aircraft efficiency and safety. Traditionally, predictive models for aircraft operation have relied on complex, black-box machine learning techniques that lack transparency. This study introduces a more interpretable approach by leveraging the New Comprehensive Modular Aero-propulsion System Simulation (N-CMAPSS) dataset and combining expanding window classification, voting schemes, and spectral clustering to detect distinct flight regimes. The method applies elastic registration to align time-shifted patterns and Functional Principal Component Analysis (FPCA) to reduce dimensionality, capturing core dynamics across flight regimes. These transformed features are fed into a genetic algorithm-assisted orthogonal matching pursuit for sparse feature selection. Through evolutionary selection, crossover, and mutation, the most informative features are identified, enabling accurate predictions while maintaining transparency. This method outperforms more complex models in certain test cases, offering a balance between accuracy and interpretability that is essential for predictive maintenance and safety applications.

Knowledge Graph Completion for High-Speed Railway Turnout Switch Machine Maintenance Based on the Multi-Level KBGC Model

The incompleteness of the existing knowledge graphs in the railway domain creates information gaps, impacting their quality and effectiveness in the operation and maintenance of high-speed railway turnout switch machines. To address this, we propose a multi-layer model (KBGC) that combines KG-BERT, graph attention network (GAT), and Convolutional Embedding Network (ConvE) for knowledge graph completion in railway maintenance. KG-BERT fine-tunes a pre-trained BERT model to extract deep semantic features from entities and relationships, converting them into graph structures. GAT captures key structural relationships between nodes using an attention mechanism, producing enriched semantic and structural embeddings. Finally, ConvE reshapes and convolves these embeddings to learn complex entity interactions, enabling accurate link prediction. Extensive experiments on the HRTOM dataset, containing triplet data from high-speed railway turnout switch machines, demonstrate the model’s effectiveness, achieving an MRR of 50.8% and a Hits@10 of 60.7%. These findings show that the KBGC model significantly improves knowledge graph completion, aiding railway maintenance personnel in decision making and preventive maintenance, and providing new tools for railway maintenance applications.

Serverless Computing in the Edge-Cloud Continuum : Challenges, Opportunities, and a Novel Framework

This article explores the integration of serverless computing across the edge-cloud continuum, addressing the growing demand for low-latency, data-intensive applications in the era of the Internet of Things (IoT). We present EdgeServe, a novel framework that extends the serverless paradigm to encompass edge devices and cloud resources, overcoming the limitations of traditional cloud-centric approaches. Through comprehensive simulations and real-world case studies, including a smart city traffic management system, an industrial IoT predictive maintenance application, and a mobile augmented reality gaming platform, we evaluate the performance, scalability, and cost-effectiveness of our proposed framework. Our results demonstrate significant improvements in application response times, with latency reductions of up to 82% in time-critical functions, and an average cost reduction of 43% compared to cloud-only serverless deployments. We also observe a 28% decrease in overall energy consumption in certain scenarios. The article addresses key challenges in resource management, data consistency, adaptive function placement, and security in distributed environments. Our findings have important implications for application architects and cloud service providers, paving the way for a new generation of edge-native serverless platforms. This research contributes to the growing body of knowledge on distributed systems and offers insights into the future evolution of serverless computing in heterogeneous computing environments.

Performance of the polyurethane foam injection technique for road maintenance applications

Road infrastructure plays an important role in strengthening transportation and driving the economic advancement of countries. However, the increasing traffic volume has accelerated road deterioration, particularly at critical points like bridge-road junctions. Traditional repair methods involving demolition and reconstruction lead to extended closures and high costs. This study explores the polyurethane (PU) foam injection technique as an alternative solution, which can reduce both repair time and costs. The research evaluates the application of PU foam in various road projects across Thailand, highlighting its ability to repair pavement surfaces and structures, even in severely damaged areas. Despite its advantages, the use of PU foam faces challenges due to a lack of standardized quality control. This paper proposes a set of working guidelines for PU foam injection, aimed at key stakeholders such as the Department of Highways, the Department of Rural Roads, and the Department of Local Administration. The findings underline the importance of establishing standardized methods to ensure the long-term effectiveness of PU foam in road maintenance. Future research should focus on refining these guidelines for diverse road conditions to support the sustainable development of national transportation infrastructure

Improving the blade tip clearance measurement method based on blade tip timing: Accounting for rotor speed variations and non-uniform blade-by-blade tip clearance

The Blade Tip Clearance (BTC) measurement method based on Blade Tip Timing (BTT) suffers from poor accuracy under changing speed conditions and in the determination of blade-by-blade tip clearance. To address these challenges, this paper presents the Time of Arrival Polynomial Fitting (ToA-PF) method for achieving high-precision tip clearance measurements under variable speed conditions. Additionally, a correction method is introduced to effectively mitigate measurement errors in blade-by-blade tip clearance. Experimental validation on an academic test rig demonstrates significant improvements. The ToA-PF method enhances clearance measurement accuracy by 77% compared to traditional methods under low and rapidly changing rotational speeds. Furthermore, the correction method for blade-by-blade tip clearance notably enhances measurement accuracy, particularly for the shortest blade, by over 50%. These proposed methods show promise in enhancing the accuracy and practicality of the BTC measurement method based on BTT, offering substantial value for turbine blade health monitoring and predictive maintenance applications.

Decomposing maintenance actions into sub-tasks using natural language processing: A case study in an Italian automotive company

Industry 4.0 has led to a huge increase in data coming from machine maintenance. At the same time, advances in Natural Language Processing (NLP) and Large Language Models provide new ways to analyse this data. In our research, we use NLP to analyse maintenance work orders, and specifically the descriptions of failures and the corresponding repair actions. Many NLP studies have focused on failure descriptions for categorising them, extracting specific information about failure, or supporting failure analysis methodologies (such as FMEA). Whereas, the analysis of repair actions and its relationship with failure remains underexplored. Addressing this gap, our study makes three significant contributions. Firstly, we focused on the Italian language, which presents additional challenges due to the dominance of NLP systems that are mainly designed for English. Secondly, it proposes a method for automatically subdividing a repair action into a set of sub-tasks. Lastly, it introduces an approach that employs association rule mining to recommend sub-tasks to maintainers when addressing failures. We tested our approach with a case study from an automotive company in Italy. The case study provides insights into the current barriers faced by NLP applications in maintenance, offering a glimpse into the future opportunities for smart maintenance systems.

Hybrid group opportunistic maintenance strategy for offshore wind turbines considering maintenance resource allocation in harsh working environments

The intelligent maintenance of offshore wind turbines (OWTs) is vital for sustainable energy generation and utilization, as harsh working environments generally limit their application. As offshore wind farms expand into deeper waters, operational and maintenance costs continue to increase. Therefore, devising targeted maintenance strategies for offshore wind farms is crucial. Existing strategies often overlook the role of maintenance resources in controlling the operational and maintenance costs. To address this problem, this study proposes a hybrid group opportunistic maintenance strategy to implement resource allocation-based maintenance for OWTs under dynamic conditions. First, a failure rate model based on the life decline and system coupling is proposed to capture the evolution of failures in OWTs in harsh working environments, thereby enhancing the accuracy of fault maintenance. Subsequently, a hybrid linear and nonlinear deterioration model is proposed to identify the different deterioration trends owing to the various sensitivities of diverse components to harsh working environments, which enhances the precision of degradation maintenance. In addition, a maintenance resource allocation model is proposed to allocate resources in advance, thereby increasing resource utilization and reducing maintenance costs. A multi-objective optimization approach is proposed to maximize the mean maintenance level (MML) and minimize the per unit cycle expected maintenance cost rate (PECR) by considering harsh working environments. Finally, the nutcracker optimization algorithm is introduced into the maintenance model to generate the optimal solution in a multi-objective manner. The results show that the maintenance model is feasible and effective for practical adaptive maintenance applications of OWTs in harsh working environments in practical applications.

Application AVATAR (Accounting Virtual Report) on PT Pelindo Terminal Petikemas

This research discusses the application of asset maintenance at PT Pelindo Terminal Petikemas. This research aims to determine the asset maintenance system at PT Pelindo Terminal Petikemas, as well as to create a web-based asset maintenance application program at PT Pelindo Terminal Petikemas. The theoretical framework of this research involves information system analysis, by examining the information needed by management, related functions, documents used, as well as document flowcharts. MySql is used as a database, with the PHP programming language. The research results show that the recording process at PT Pelindo Terminal Petikemas has been computerized but still uses Microsoft Excel. The weakness of this method is that it is difficult to look back at archived data for the previous year because you need to search for the file first and there is a risk of losing the file. Therefore, the application called AVATAR (Accounting Virtual Report) in this research is expected to help produce asset maintenance reports more effectively and efficiently, thus providing benefits for the company.

Railway condition information modelling in a BIM framework for maintenance applications

ABSTRACT Railway maintenance management involves various interventions that, considering safety and use needs, require an integrated approach for supporting decisions. Defining an infrastructure digital twin can represent an efficient basis to implement management processes. BIM, offering a simplified and realistic digitalized model, may ensure immediate benefits for asset administrators in maintenance management and planning. It would permit to properly maintain full control of infrastructure degradation, to optimize investments and ensure adequate quality levels. In this context, a preliminary procedural framework for modelling railway condition information in an I-BIM environment is proposed, using specific survey data acquired by scheduled diagnostic trains. The authors developed customized routines and scripts to elaborate rail track geometry attributes (with real-time checks of the dynamic effects on the track) and simplify asset quality evaluation for the administrators. This may suggest a novel vision and framework to support maintenance management procedures in future developments of the overall approach.

Enhancing waste asphalt durability through cold recycling and additive integration

The longevity of waste asphalt can be considerably improved through cold recycling techniques combined with various additives. This research investigates the cold regeneration of aged asphalt concrete using Reclaimed Asphalt Pavement (RAP), Portland cement, cationic bitumen emulsion, and additional aggregates. The primary goal is to evaluate the performance enhancements in terms of average density, compressive strength, water resistance, and swelling across different mix compositions. Three distinct mixtures were formulated and assessed. Mix No. 1, composed solely of RAP, showed the lowest average density and highest swelling, indicating poor performance due to the lack of binding agents. Mix No. 2, which incorporated RAP, Portland cement, and water, exhibited the highest density and compressive strength, highlighting the crucial role of Portland cement in improving structural integrity. Mix No. 3, a more complex mixture including RAP, aggregates, Portland cement, water, and bitumen emulsion, displayed balanced properties with enhanced moisture resistance and reduced swelling. The experimental findings emphasize the effectiveness of adding Portland cement and bitumen emulsion to improve the mechanical and durability characteristics of recycled asphalt mixtures. Specifically, Mix No. 2 and Mix No. 3 demonstrated significant performance improvements, making them suitable for road maintenance applications. This study advocates for the widespread use of cold recycling methods with additive integration to achieve sustainable and cost-effective pavement restoration solutions.

Design and Implementation of an Interactive Embedded System as a Low-Cost Remotely Operated Vehicle for Underwater Applications

The underwater environment is harsh and challenging for human life, prompting companies and researchers to develop advanced technologies for exploration. Building on previous work that applied a CNN-based method for underwater object classification, this paper focuses on the design and implementation of an interactive embedded system for a compact Remotely Operated Vehicle (ROV) with specific dimensions and weight. The primary goal is to capture real-time underwater video using remote control communication via Ethernet. The ROV is powered by five brushless motors controlled by a smart PID controller. Precise pulse-width modulation signals enhance stability during movements along three axes, enabling high-resolution video capture. The system utilizes the Raspberry Pi 3's computing power for motion control, positioning, temperature monitoring, and video acquisition. Experimental results demonstrate the system's capability to process 42 frames per second. A user-friendly graphical interface allows for remote ROV control across various operating systems. With a depth rating of 100 meters and speed of 0.148 m/s. This ROV surpasses human divers' limitations and holds significant potential for applications in surveillance, operations, maintenance, and measurement tasks underwater. The proposed ROV contributes to the advancement of underwater exploration technology by combining high performance with cost-effectiveness.

Beyond seen faults: Zero-shot diagnosis of power circuit breakers using symptom description transfer

Power circuit breakers (CBs) are vital for the control and protection of power systems, yet diagnosing their faults accurately remains a challenge due to the diversity of fault types and the complexity of their structures. Traditional data-driven methods, although effective, require extensive labeled data for each fault class, limiting their applicability in real-world scenarios where many faults are unseen. This paper addresses these limitations by introducing symptom description transfer-based zero-shot fault diagnosis (SDT-ZSFD), a method that leverages zero-shot learning for fault diagnosis. Our approach constructs a fault symptom description (FSD) framework, which embeds a fault symptom layer between the feature layer and the label layer to facilitate knowledge transfer from seen to unseen fault classes. The method utilizes current and acceleration signals collected during CB operation to extract features. By applying sparse principal component analysis to these signals, we derive high-quality features that are mapped to the FSD framework, enabling effective zero-shot learning. Our method achieves a satisfactory recognition rate by accurately diagnosing unseen faults based on these symptoms. This approach not only overcomes the data scarcity problem but also holds potential for practical applications in power system maintenance. The SDT-ZSFD method offers a reliable solution for CB fault diagnosis and provides a foundation for future improvements in symptom-based zero-shot diagnostic mechanisms and algorithmic robustness.

An Efficient Computational System For Defect Prediction through Neural Network And Bio-inspired Algorithms

Detecting and locating damage is essential in maintaining structural integrity. While Artificial Neural Networks (ANNs) are effective for this purpose, their performance can be significantly improved through advanced optimization techniques. This study introduces a novel approach using the Grasshopper Optimization Algorithm (GOA) to enhance ANN capabilities for predicting defect aluminum plates. The methodology begins by deriving input parameters from natural frequencies, with defect locations as the output. A Finite Element Model (FEM) is used to simulate data by varying defect locations, creating a comprehensive dataset. To validate this approach, experimental data from vibration analyses of plates with different defect locations is collected. We then compare the performance of our GOA-optimized ANN against other metaheuristic algorithms, such as Cuckoo Search Algorithm (CSA), Bat Algorithm (BA), and Firefly Algorithm (FA). Notably, CSA's performance is slightly close to GOA. The results show that our GOA-based method outperforms these traditional algorithms, demonstrating superior accuracy in damage prediction. This advancement holds significant potential for applications in structural integrity monitoring and maintenance.

A scientometric analysis and critical review of digital twin applications in project operation and maintenance

PurposeRecent emerging information technologies like digital twin (DT) provide new concepts and transform information management processes in the architecture, engineering and construction (AEC) industry. Although numerous articles are pertinent to DT applications, existing research areas and potential future directions related to the state-of-the-art DT in project operation and maintenance (O&M) are yet to be studied. Therefore, this paper aims to review the state-of-the-art research on DT applications in project O&M.Design/methodology/approachThe current review adopted four methodological steps, including literature search, literature selection, science mapping analysis and qualitative discussion to gain a deeper understanding of DT in project O&M. The impact and contribution of keywords and documents were examined from a total of 444 journal articles retrieved from the Scopus database.FindingsFive mainstream research topics were identified, including (1) DT-based artificial intelligence technology for project O&M, (2) DT-enabled smart city and sustainability, (3) DT applications for project asset management, (4) Blockchain-integrated DT for project O&M and (5) DT for advanced project management. Subsequently, research gaps and future research directions were proposed.Originality/valueThis study intends to raise awareness of future research by summarizing the current DT development phases and their impact on DT implementation in project O&M among researchers and practitioners.

Specifics of Data Collection and Data Processing during Formation of RailVista Dataset for Machine Learning- and Deep Learning-Based Applications.

This paper presents the methodology and outcomes of creating the Rail Vista dataset, designed for detecting defects on railway tracks using machine and deep learning techniques. The dataset comprises 200,000 high-resolution images categorized into 19 distinct classes covering various railway infrastructure defects. The data collection involved a meticulous process including complex image capture methods, distortion techniques for data enrichment, and secure storage in a data warehouse using efficient binary file formats. This structured dataset facilitates effective training of machine/deep learning models, enhancing automated defect detection systems in railway safety and maintenance applications. The study underscores the critical role of high-quality datasets in advancing machine learning applications within the railway domain, highlighting future prospects for improving safety and reliability through automated recognition technologies.

Development of an Optimization-Based Budget Allocation Model for Seismic Strengthening Based on Seismic Risk Assessment

This study presents a technology used for the prediction of economic losses to facilities in a given area during an earthquake, thereby enabling the efficient application of performance-based maintenance and seismic strengthening. We also propose an algorithm for the establishment of a reinforcement plan that minimizes earthquake-induced economic losses within a constrained budget. The algorithm incorporates fragility functions from prior research and utilizes an optimization technique for budget allocation, leveraging the target damage ratio concept and constrained optimization. Based on the fragility curve, the probability of occurrence for each damage state for a specific PGA value and the damage rate for each damage state are calculated. From these values, the expected damage ratio (EDR) is estimated. An optimization-based budget allocation algorithm is developed to find the elements that would result in the lowest damage rate for a limited cost. To validate the applicability of the model, we created a hypothetical city with a 30 km × 30 km area containing bridges, embankments, and buildings. The estimated pre- and post-reinforcement damage was assessed in two earthquake scenarios, allowing us to test the effectiveness of the optimization-based budget allocation model in reducing damage. These results suggest that the proposed model offers a viable strategy for efficient seismic strengthening within budgetary constraints.