Equipment Failure Research Articles

Quantifying the Impact of Sustained Acceleration on Critical Care Transport Medical Equipment

ABSTRACT Introduction Military and commercial stakeholders are investing to explore the use of hypersonic aircraft and orbital spacecraft to transport cargo, medical supplies, passengers, and casualties. These vehicle platforms require periods of sustained acceleration, but to date, these dynamic forces have not been comprehensively considered in the environment of critical care patient movement because injured patients and advanced aeromedical evacuation (AE) equipment are rarely subjected to these conditions. While military AE equipment does undergo crash hazard acceleration testing, equipment functionality during or after sustained acceleration remains to be evaluated. This study was performed to fill that knowledge gap. Materials and Methods AE equipment currently used by the U.S. Air Force and Critical Care Air Transport Teams (ZOLL EMV+ 731 ventilator and ZOLL Propaq MD cardiac monitor) was subjected to low (2.5 g), moderate (4.5 g), and variable acceleration (1.5–4.5 g) for 3-minute periods at the KBR Brooks Centrifuge. AE equipment was tested for functionality in 3 different orientations (gX, gY, and gZ). Predetermined variations were made in equipment input settings to ensure each equipment item would function across mission-relevant conditions (differing ventilator tidal volumes, differing cardiac monitor arterial pressure inputs, etc.). AE was evaluated for accuracy compared to controlled inputs, alarm conditions, and equipment failure. Results The EMV+ 731 ventilator and Propaq MD cardiac monitor had no equipment failures during testing. The ventilator had clinically negligible variations in tidal volume, peak pressure, and fraction of inspired oxygen during acceleration. At the highest tidal volume tested (480 mL), the ventilator had elevated peak pressure results. However, we believe this was due to limitations in test-lung resistance and was not related to a ventilator fault. Mild effects of sensor orientation were recorded in the Propaq MD blood pressure results; for example, gX and gY differed by 5.9 ± 1.21 mmHg at 75 mmHg input (P < .01) and 6.45 ± 1.229 mmHg with 150 mmHg input (P < .001). Conclusions The EMV+ 731 ventilator and Propaq MD cardiac monitor had reliable clinical performance despite sustained acceleration. This knowledge will facilitate immediate follow-on experimentation with advanced models of combat injury during simulated medical evacuation in sustained acceleration environments.

A System Dynamics Stability Model for Discrete Production Ramp-Up

Manufacturing companies are increasingly challenged to deliver customizable products with shorter time to market and higher quality while adhering to sustainability requirements. To meet these challenges, the frequency and importance of production ramp-ups will increase in the future. However, most ramp-ups still fail to meet targets due to unpredictable equipment failures, operator errors, and system complexity. We propose a system dynamics model that captures the unique dynamics of ramp-up phases by integrating stability and disturbance factors that influence the key performance indicators overall equipment effectiveness, process capability, and production output. A systematic literature review informed the identification of stability factors, which were validated through expert interviews in the automotive industry. Our system dynamic simulation results indicate that control factors realistically influence production system behaviour during different ramp-up phases. Despite some limitations regarding the effects of maintenance personnel and engineering changes on key performance indicators, our model effectively simulates realistic ramp-up behaviour. The findings highlight the need for tailored models that consider specific ramp-up contexts and emphasize the importance of data acquisition for enhanced performance prognosis in future research.

Noise Reduction in CWRU Data Using DAE and Classification with ViT

With the Fourth Industrial Revolution unfolding worldwide, technologies including the Internet of Things, sensors, and artificial intelligence are undergoing rapid development. These technological advancements have played a significant role in the dramatic growth of the predictive maintenance market for mechanical equipment, prompting active research on noise removal techniques and classification algorithms for the accurate determination of the causes of equipment failure. In this study, time series data were preprocessed using the denoising autoencoder technique, a deep learning-based noise removal method, to improve the accuracy of failure classification from mechanical equipment data. To convert the preprocessed time series data into frequency components, the short-time Fourier transform technique was employed. The fault types of mechanical equipment were classified using the vision transformer (ViT) technique, a deep learning technique that has been actively used in recent image analysis research. Additionally, the classification performance of the ViT-based technique for vibration time series data was comparatively validated against existing classification algorithms. The accuracy of failure classification was the highest when the data, preprocessed using a Denoising Autoencoder (DAE), were classified by a Vision Transformer (ViT).

IMPROVING THE EFFICIENCY OF ENERGY COMPANIES BY OPTIMISING THE DISTRIBUTION OF COSTS FOR MAINTENANCE AND REPAIR OF ELECTRICAL EQUIPMENT

At present, there is a high emergencies probability in the electric power system as an electrical equipment failures result caused primarily by the objectively existing ageing and its service life significant exhaustion. Emergencies arising from electrical equipment failures lead to the dynamic stability violation of power systems and power system load nodes by voltage, cascading accidents and, as a result, technological processes disruption of consumer enterprises, which is accompanied by significant losses. One of the ways to prevent sudden failures or minimise the risk of their occurrence is to ensure rational and efficient operation management and reduce the operating equipment cost without risking a decrease in the power system reliability. To solve the optimal distribution problems of power company operating costs, taking into account the objectively existing limiting factors, a generalised balance sheet model of cost allocation by electrical equipment groups and units is proposed, which is a total costs function of an operating factors number. Mathematical models for calculating the electrical equipment share participation coefficients in the costs formation for its repair and maintenance for each of the identified factors have been obtained. A linguistic mathematical model for determining generalised estimates of the electrical equipment units and groups significance degree in the repair and maintenance costs distribution formation, taking into account the main operational factors, has been proposed. An algorithm and software for modelling the modes of the electric power system with an electrical connections scheme of electric networks with a 35-330 kV voltage have been developed to assess the operation risk index and power company repair and maintenance costs distribution. The estimated power company repair and maintenance costs distribution according to the current technical equipment characteristics and power system subsystem modes was performed.

Analysis of Hazard Risk Understanding Using "Safety-Uno" Education Media for Batik Workers in Batik X and the General Public

Batik is the art of drawing on cloth using canting with small ends and batik wax. The process of making batik can cause work accidents and occupational diseases that occur because of equipment failure and human errors that do work not according to procedures. However, in the informal batik industry, there still needs to be more knowledge and awareness related to potential hazards and controls that must be carried out. Therefore, it is necessary to conduct education to increase workers' understanding and create a safe and comfortable work environment. This research method is a quantitative study using the Wilcoxon Test to compare observations before and after treatment and determine the effectiveness of treatment. Based on the observations and analysis that have been carried out, it can be concluded that there is a significant relationship between the use of safety cards.

Quantitative Risk Assessment of Hydrotreated Vegetable Oil at an Oil and Gas Company

Introduction: An oil and gas refinery operates various equipment with specific functions for different processes. Each piece of equipment has potential hazards that can damage the equipment and injure or kill workers. This study focuses on the hydrotreated vegetable oil (HVO) export facility from the jetty loading area at an oil and gas company that processes flammable liquid using various equipment. Methods: The HAZOP method determined the hazardous spots, and the probability of each equipment failure corresponding to the system was also determined using fault tree analysis (FTA). Furthermore, every event tree analysis (ETA) output probability was also determined. The probability and radius of pool fire varied for different leak hole scenarios. The final steps are individual risk per annum and potential loss of life to measure the risk level of the system. Results: Based on HAZOP deviation scenarios, every operating equipment can potentially cause a pool fire. In FTA, scenarios were developed based on different leakage hole sizes, ranging from 1-3 mm, 3-10 mm, 10-50 mm, 150 mm, and >150 mm. The results indicated that leakage could occur across all operating equipment. Similarly, the ETA applied the same bore size scenarios. The consequence analysis yielded a worst-case outcome of pool fire and a best-case outcome of un-ignited fluid release. Subsequently, the pool fire output was modeled using ALOHA, which resulted in three heat flux zones: the red zone (10 kW/m²), the orange zone (5 kW/m²), and the yellow zone (2 kW/m²). Smaller leak holes had a higher probability but smaller pool fire radius. The initial risk of the export facility was unacceptable. Furthermore, insufficient safeguards contribute significantly to the resulting high-risk level. Two mitigations were implemented: adding safeguards and reducing worker hours. Conclusion: The final results showed that for every piece of equipment, the overall risk of the export facility became acceptable after mitigation..

Evaluating Pilot Mental Workload Using fNIRS-Based Functional Connectivity Features with a Deep Residual Shrinkage Network Under Emergency Flight Scenarios

Excessive mental workload can lead to less remaining resources for pilots to perform concurrent tasks during emergency flights, affecting aviation safety. Based on a flight simulator, this study investigated 25 cadet pilots using functional near-infrared spectroscopy (fNIRS) and subjective ratings to assess their mental workload under three subtasks with different equipment failures. fNIRS data included oxyhemoglobin, deoxyhemoglobin, and total hemoglobin signals, yielding 10545 functional connectivity (FC) features from four brain regions: prefrontal, right motor, left motor, and occipital cortexes. A deep residual shrinkage network classified mental workload levels, outperforming convolutional neural network and random forest models with 89.58% accuracy after feature selection employing an interpretable machine learning algorithm. The results suggest that brain FC from three hemoglobin signals could be used to differentiate the three different levels of pilot mental workload. This study could contribute to improving pilot training and supporting the development of pilots’ competencies during emergency scenarios.

Short-term prediction of power outages in electrical distribution networks

Predictive maintenance and reliability engineering are critical in industrial settings to enhance equipment performance and minimize unplanned downtime. This research, conducted within the machine learning framework, presents innovative solutions to the challenging problem of equipment failure prediction. The study creatively utilizes extensive datasets, including equipment records, weather conditions, and maintenance logs, to develop robust predictive models. Two distinct machine learning models are established for equipment and cables/lines, addressing the intricacies of class imbalances and missing data attributes. Model refinement, feature engineering, and interdisciplinary collaboration enhance predictive accuracy, precision, and recall. Notably, this research highlights the creative application of engineering knowledge and data science techniques, reasoning about complex equipment systems, and the importance of problem decomposition. The outcomes underscore the potential for real-time predictive maintenance in industrial contexts, offering substantial cost savings and improved equipment reliability. This research contributes to the evolving field of predictive maintenance and paves the way for future innovations in reliability engineering.

Building a Culture of Maintenance: Strategies for Addressing Office Machine Challenges in Polytechnics

This research study was undertaken to investigate the various challenges faced by staff in maintaining office machines, which are critical to the smooth operation of modern workplaces. The study involved a thorough review of existing literature, encompassing both published and unpublished works, to examine the wide array of issues related to the maintenance of office equipment. To gather empirical data, a structured questionnaire comprising 15 items was developed to address three specific research questions. This questionnaire was administered to a sample of 35 staff members from selected polytechnics in the North Western Zone. The data collected from the responses were systematically analysed using the mean method, following a 5-point Likert-type rating scale. The findings were organised and presented in well-structured tables to facilitate understanding and interpretation. The analysis highlighted that one of the most effective measures staff can adopt to prevent breakdowns of office machines is to have a comprehensive understanding of contemporary office equipment and their maintenance needs. Furthermore, the results revealed that a significant contributing factor to the frequent breakdown of these machines is the lack of proactive organisational decisions and comprehensive maintenance policies. Additionally, a marked disinterest in identifying the root causes of equipment failures further exacerbates the challenges faced by staff. Based on these findings, this study recommends that organisations implement a series of conferences, workshops, and training sessions focused on maintenance best practices. Such initiatives would serve to enhance awareness among staff about effective strategies for prolonging the lifespan of office machines. Furthermore, by fostering a culture of proactive maintenance, organisations can significantly reduce their maintenance costs and improve operational efficiency.

A method for the core accident chain based on fuzzy-DEMATEL-ISM: An application to aluminium production explosion

The occurrence of accidents is often influenced by many factors, which result in the complexity of the causes of accidents. Therefore, clarifying the accident chain of the explosion accident is of great significance for the daily management of safety. In this study, a novel method of the core accident chain based on the traditional fuzzy-DEMATEL-ISM method was proposed. This method can clarify the importance of each accident chain and rank them accordingly by calculating their weighted centralities. Then, the explosion accident caused by the contact of molten aluminium and water, which has caused heavy casualties and property losses, is analyzed systematically. In order to clarify the impact relationships and hierarchical structure of the various factors in the complex explosion accident systems, and establish the accident system models, twenty-two factors were summarized, and the evolution process of explosion accident were obtained by the fuzzy-DEMATEL-ISM approach. The results reveal that equipment damage or failure has the most direct impact to the explosion accidents. The core accident chain that national policy (aluminium market demand)→incomplete safety management system→disordered on-site management→insufficient lighting→physical discomfort→improper operation→equipment damage or failure is obtained, which has the most significant influence on the occurrence of explosion accidents. The result analysis is employed to give some advice. This research is helpful for safety practitioners to develop accident prevention strategies and make aluminum production safer.

Trajectory reconstruction of buoys using Qingdao harbour data: Filling missing buoy trajectories using LSTM-Attention model

In maritime operations, instances occur where vessels flee after colliding with buoys. Automatic Identification System (AIS) data collected from these buoys play a vital role in identifying the responsible party and determining liability. However, data from buoys can be lost in rough sea conditions and equipment failure, which makes it difficult to identify accidents where ships collide with buoys. To address this issue, this study proposes a model that combines Long Short-Term Memory Network (LSTM) with Multiple Attention mechanisms to reconstruct buoy trajectories with high precision. In this paper, the buoy trajectory reconstruction test is carried out using Qingdao Harbour Buoy 304 a case study, and the validation results show that the mean squared error (MSE) of this model is 12.15. In addition, the LSTM-Attention model shows a significant improvement in all the metrics compared with other models: compared with the Random sampling model, the mean absolute error (MAE) is improved by 75.15%, and the root mean squared error (RMSE) by 71.02%; MSE by 77.31%, MAE by 36.70%, and RMSE by 52.36% compared to the Cubic spline model; and MSE by 49.92%, MAE by 27.35%, and RMSE by 29.52% compared to the Hermite model. These results show that the LSTM-Attention model significantly improves the accuracy and reliability of trajectory reconstruction.

Deep one-class classification model assisted by radius constraint for anomaly detection of industrial control systems

Anomaly detection of industrial control systems (ICS) based on sensor data analytic is of utmost importance because ICS may suffer from various attacks leading to anomaly behaviors and even equipment failures. As an emerging deep learning technique, deep support vector data description (DeSVDD) has been successfully applied to ICS anomaly detection. Its advantage lies in only requiring normal data to train one-class classifier, which is suitable for actual industrial scenarios with extremely data imbalance of abundant normal data and scare abnormal data. However, this also results in its shortcoming of ignoring the valuable classification information hidden in the abnormal data. In order to overcome this issue, this paper proposes an improved DeSVDD method, called radius constraint DeSVDD (RC-DeSVDD). The proposed model constructs an anomaly detection model framework of abnormal data assisted DeSVDD, where a radius constraint is designed by considering the difference between normal and abnormal data. Further, considering the limited amount of abnormal data, a bi-directional generative adversarial network (BiGAN) is introduced to generate abnormal data. Experimental results on three benchmark datasets demonstrate the superiority of the proposed RC-DeSVDD anomaly detection method.

Highly electrically conductive polyether composites with modified graphene

Abstract With the continuous improvement in the voltage, power, and capacity levels of high-voltage transmission and substation equipment, the problems of power loss and equipment failure caused by the abnormal heating of electrical contact parts are becoming increasingly severe. In the present study, to address this problem, graphite was exfoliated into thin layers of graphene using liquid-phase mechanical exfoliation, ultrasonic dispersion, and spray-drying techniques and incorporated into polyether composites to increase its electrical conductivity. The effects of the graphene content on the electrical conductivity, high-temperature resistance, wear reduction, and antiwear properties of the polyether composites were investigated. The results indicated that when 4 wt% graphene was added, the high-temperature resistance of the graphene–polyether composite (GPC) increased to 330 °C, and the volume resistivity decreased to 6.5 × 103 Ω·cm. Moreover, the contact-resistance coefficient of the GPC was reduced to 0.87 and 0.73 after it was coated on Cu and Al rows, respectively, which significantly increased the electrical conductivity of the electrical contact area. The most significant improvements in friction-reduction and antiwear properties were obtained for the polyether composites from this formulation. GPC has excellent electrical conductivity, high-temperature resistance, wear reduction, and antiwear properties and thus can substantially improve the quality of electrical connections when applied to electrical contact tips.

Reducing the risk of power equipment failure when using information from information measurement system of vibration diagnosis rotating units of auxiliary engines of power plant

Reducing the risk of power equipment failure when using information from information measurement system of vibration diagnosis rotating units of auxiliary engines of power plant

Novel imbalanced multi-class fault diagnosis method using transfer learning and oversampling strategies-based multi-layer support vector machines (ML-SVMs)

For health monitoring and fault diagnosis of critical mechanical system components, historical data related to equipment failures are often limited and exhibit varying imbalanced multi-class characteristics (e.g., with noisy and time-series data). Moreover, fault diagnosis frameworks based on traditional resampling algorithms (e.g., SMOTE) mostly heavily rely on manual feature extraction, making them difficult to adapt to diverse working conditions or objects. To address these challenges, we propose a novel end-to-end imbalanced multi-class fault diagnosis architecture using transfer learning and oversampling strategies-based multi-layer support vector machines (ML-SVMs). ML-SVMs utilize a VGG-based deep migration feature extraction method to extract features from original time-domain vibration signals, employing natural source domain weights to reduce dependence on human experience and sample size. Then, ML-SVMs introduce ISCOTE (i.e., the first and second layers of ML-SVMs), an improved version of the sample-characteristic over-sampling technique (SCOTE). ISCOTE generates more effective and reasonable samples for each fault class through a scaling factor and iterative optimization mechanism, whether in noisy feature spaces with fuzzy boundaries or in clear boundary feature spaces. Finally, in the third layer of ML-SVMs, multi-class SVMs (e.g., LS-SVMs and standard SVMs) are utilized to train balanced feature samples and derive classification models with strong generalization ability. The effectiveness of ML-SVMs is demonstrated through 16 fault diagnosis instances using CWRU and IMS bearing data, PHM 2010 and TTWD tool wear data. Results indicate that ML-SVMs outperform 8 well-known oversampling-based algorithms in fault diagnosis recognition rates and algorithm robustness. It has offered a feasible architecture for multi-class imbalanced fault scenarios with limited data and multiple adverse features.

Strengthening environmental safety in oil and gas operations: Optimizing health, safety, and environmental (HSE) protocols

The oil and gas industry faces significant environmental challenges, including spills, emissions, and equipment failures, which pose serious risks to ecosystems and human health. This review explores strategies for optimizing Health, Safety, and Environmental (HSE) protocols in oil and gas operations, emphasizing technological advancements, proactive safety measures, and industry-regulator collaboration. It examines how real-time monitoring systems, predictive analytics, and automation improve risk management, while best practices such as comprehensive risk assessments and safety training mitigate potential hazards. The paper also highlights the need for flexible regulatory frameworks and enhanced transparency through data sharing. The findings underscore the importance of refining HSE protocols to ensure environmental safety and operational resilience. Recommendations include expanding the use of advanced technologies, strengthening safety cultures, and fostering closer cooperation between the industry and regulatory bodies. These strategies aim to reduce environmental risks, improve compliance, and enhance sustainability in the sector.

AI-enhanced predictive maintenance systems for critical infrastructure: Cloud-native architectures approach

Critical infrastructure (CI), such as power grids, transportation systems, and telecommunications networks, is becoming increasingly complex, requiring sophisticated maintenance strategies and procedures to guarantee optimal performance and system durability. This paper examines the transformational potential of AI-driven predictive maintenance systems, highlighting their ability to prevent system failures, minimize downtime, and enhance resource efficiency. Integrating machine learning algorithms with real-time data analytics allows predictive maintenance frameworks to accurately foresee equipment failures, facilitating timely interventions that reduce the risk of catastrophic infrastructure breakdowns. This study primarily examines the development of cloud-native architectures, which include containers, microservices, and orchestration tools like Kubernetes, to facilitate the scalability, flexibility, and resilience required for contemporary CI maintenance systems. These designs facilitate the seamless integration of predictive maintenance solutions across geographically dispersed infrastructure, enabling effective administration of extensive datasets produced by Internet of Things (IoT) sensors, operational logs, and edge computing nodes. The document examines the essential function of intelligent data orchestration in facilitating the prompt gathering, processing, and analysis of operational data, which is vital for AI models to provide precise predictions. The amalgamation of AI-driven predictive maintenance with 5G and forthcoming 6G networks is poised to transform real-time system monitoring, diminishing latency and enhancing decision-making efficacy. Utilizing AI and cloud-native technologies substantially enhances system reliability, cost-effectiveness, and comprehensive infrastructure optimization. This article thoroughly analyses how AI, cloud-native platforms, and intelligent data orchestration may be utilized to tackle the changing maintenance issues of critical infrastructure by examining real-world case studies from sectors like power grids, telecommunications, and transportation. Integrating AI, cloud computing, and IoT in predictive maintenance improves system reliability and prepares critical infrastructure for future autonomous management and optimization developments. The study finishes by discussing new trends, such as the integration of digital twins and the synergies between AI and cloud-native solutions, which will enhance predictive maintenance capabilities.

Distributionally robust CVaR optimization for refinery integrated production–maintenance scheduling under uncertainty

In the petroleum refining industry, efficient production planning and maintenance scheduling are crucial for economic performance and operational efficiency. Moreover, the production processes face significant uncertainties stemming from market fluctuations and equipment failures. However, traditional optimization methods often treat production and maintenance independently and neglect the risk management associated with uncertainties in the production process, leading to unreliable plans and suboptimal execution. To address these issues, this paper proposes an innovative data-driven distributionally robust conditional value-at-risk (DRCVaR) method to tackle the integrated production–maintenance optimization problem under crude oil price uncertainty. By constructing confidence sets with L2 norm constraints based on historical data, our approach directly links the model’s conservatism to the amount of available data, effectively managing risk. In addition, we propose robust linear transformation to simplify the min–max nonlinear problem into a conic constraint problem, enhancing solution efficiency and ensuring better operational stability. Refinery case studies demonstrate that the proposed DRCVaR consistently achieves a practical and acceptable solution, significantly outperforming state-of-the-art approaches.

Hydrogen Leakage Risks and Mitigation Measures in Large Underground Garages

Abstract Hydrogen fuel cell vehicles, characterized by zero emissions, pollution-free operation, and high efficiency, have emerged as a key focus in the development of the global automotive industry. The operating pressure for onboard hydrogen storage tanks commonly ranges from 30 to 70 MPa. Due to hydrogen's wide combustion and explosion concentration range and its exceptionally rapid combustion rate, there is a high risk of explosions and other accidents once equipment failure happens during storage and transportation. The research presented in this paper focuses on the analysis of hydrogen leakage from storage tanks in an underground garage using FLUENT simulations. The findings reveal that released hydrogen forms a jet from the storage tank under high pressure, dispersing along the ceiling upon reaching it and accumulating at the edges and corners. Moreover, larger leakage ports on the storage tank result in a greater mass flow of hydrogen, leading to an expanded diffusion range of the hydrogen cloud and decreased local concentration. To mitigate the risk of hydrogen combustion and explosion within the garage, this study introduces 16 extraction vents on the garage ceiling and 6 natural vents on its sides. The validation of the proposed hydrogen risk mitigation measures demonstrates their effectiveness in reducing the concentration and range of flammable clouds within the garage, especially when dealing with larger leakage ports.

Towards resilience in Industry 5.0: application of system dynamics in matrix-structured manufacturing system

In the face of increasing global disruptions, resilient manufacturing systems are critical to ensuring operational stability, a concept central to the Industry 5.0 vision. The Matrix-Structured Manufacturing System (MMS) is a key example of such resilience, designed to maintain functionality and recover quickly under major disruptions. This paper introduces a novel System Dynamics (SD)-based modelling and simulation framework to quantitively assess the performance indicators and resilience of MMS. By simulating key disruption scenarios, including partial equipment failures and order fluctuations with resource constraints, this study evaluates the system’s response and recovery capabilities using a Comprehensive Resilience Index (CRI). Unlike conventional approaches, the proposed SD model captures the dynamic interactions across various production configurations and simulates the operational logic governing them, which can also be applied to other production systems. The simulation results demonstrate the potential of SD to support decision-making by providing insights into the effectiveness of reconfiguration strategies and resource management. This work aims at offering robust methods for quantitative resilience assessment and strategic planning towards the practical implementation of resilient manufacturing in Industry 5.0 envisions.