Failure Prediction Research Articles

A method of quantitative chemiluminescence immunoassay for the concentration of Growth differentiation factor-15

Growth differentiation factor-15 (GDF-15), a member of the transforming growth factor (TGF-β) superfamily, and is expressed and secreted in response to inflammation, oxidative stress and hypoxia. It has been shown in several studies to be a predictor of heart failure. However, the only kits available on the market are ELISA kits, which are costly and error-prone and are not conducive for clinical use. Here, we developed a chemiluminescence kit which optimized the reaction conditions and the reaction time was reduced to 10 min. We further proved that it can be used to measure GDF-15 in serum or plasma accurately and fastly, and provide additional information for the diagnosis of heart failure disease. Methodological comparison and clinical study verified this method is a reliable, economical and highly automated blood test method.•All necessary steps and the reagents needed are provided.•Reliability of the chemiluminescence immunoassay was verified by analyzing serum GDF-15 levels from different groups.•GDF-15 can provide clinicians with reliable prediction and disease assessment of heart failure.

Numerical simulation and experimental testing for static failure prediction in additively manufactured below-knee prosthetic sockets.

The socket of a transtibial prosthesis is a structural part customized to a patient's amputated residual lower limb. The free-form geometry of the socket can be suitable for additive manufacturing (AM) to save time and cost. However, the mechanical fracture of additively manufactured lower limb prostheses is not yet fully understood. A novel experimental method and numerical approach by finite element method (FEM) to test the strength and fracture behavior of a lower limb prosthetic socket of acrylonitrile butadiene styrene (ABS), reverse-engineered using computer-aided design (CAD) from the actual amputee's residual limb and manufactured using fused filament fabrication (FFF) are proposed in the present work. The mechanical behavior, von Mises stress distribution, and the damage status of layered AM sockets of different thicknesses were simulated by FEM using Hashin's transversely isotropic mechanical damage model, initially developed for composite materials. The experimental work showed that the fracture failure initiated at the corner of the lobe in the 4 mm thickness socket at a failure load of 918.5 N. The FEM results predicted this failure load to be 896.6 N, with only a 2.45% error as compared to the experiment. The failure loads predicted by FEM in the sockets with thicknesses of 3, 5, and 6 mm were 618.1, 1008.6, and 1105.2 N, respectively. The present work provides a dependable method for testing a below-knee prosthetic socket against static failure and arriving at a factor-of-safety (FoS) based socket thickness selection for any amputee.

Study of single point incremental forming limits of Al 1050/Mg-AZ31B two-layer sheets fabricated by roll bonding technique: Finite element simulation and experiment

In this work, for the first time, the two-layer Aluminum (Al 1050) and magnesium (AZ31B) sheet were used in the single point incremental forming (SPIF) process. Due to the high strength-to-weight ratio of the Al 1050/Mg-AZ31B sample, today it is used in significant industries such as automotive and aerospace. The finite element simulation process was carried out using the Abaqus Software considering five approaches to determine forming limits. These approaches were forming limit diagram criterion (FLDcrt), effective strain rate (ESR), second derivation of thinning (SDT), Major strain rate (MSR) and Thickness strain rate (TSR) methods. The results showed that the failure prediction in the FLDcrt method was more accurate than other methods. The thickness distribution process was carried out both by simulation and experiment to validate the simulation results. The results show a good agreement between simulation and experimental results. Finally, the results obtained from the SPIF process were compared with the FLD results of the conventional hemispherical punch test, and it was observed that the FLD was higher in the SPIF process. Moreover, the layer arrangement, which affects formability, was investigated. The results showed that the FLD with Mg-AZ31B/Al 1050 arrangement fails in higher strains compared to Al 1050/Mg-AZ31B. After validating the simulation results, the parameter effect of the step-down and tool diameter on the two-layer formability was investigated by the design of experiment. Considering the greater importance of the step-down and tool diameter effect on the created stresses and forming time compared to other parameters, their effect will be investigated. The results showed an optimal condition for the tool diameter, but an opposite relationship with formability was obtained for the step-down.

Experimental and computational failure analysis of hydrogen embrittled steel cords in a reinforced thermoplastic composite pipe

Reinforced thermoplastic pipes (RTPs) have attracted considerable attention lately in the oil and gas sector as potential alternative to carbon steel pipes. However, some RTP systems have been experiencing failures as a result of deficient design and inadequate operation practices. This study presents a comprehensive assessment of the failure of a steel-reinforced flexible thermoplastic composite layered pipe in a sour environment characterized by hydrogen sulfide (H2S) exposure. Utilizing scanning electron microscopy (SEM) and electron dispersive spectroscopy (EDS), the investigation conclusively attributes the failure of the RTP to hydrogen embrittlement of the carbon steel cords and wires in the reinforcement layer. Finite element models at both meso-scale and full RTP scale were developed, incorporating a novel damage initiation and propagation criterion for failure prediction of embrittled and non-embrittled (e.g. ductile) reinforcement steel cords. The numerical models consistently predicted a reduced burst pressures of the hydrogen embrittled case of the RTP, with a close match with field observations. They also predicted failure modes that closely matched the SEM findings, hence highlighting the unsuitability of the existing steel cords in RTPs for use in sour service environments. Additionally, the presence of moisture in the annular space was identified as an exacerbating factor. This investigation provides important insights for the future design and operation of RTPs susceptible to hydrogen embrittlement, paving the way for improved structural integrity and safety measures. Further research avenues may involve modeling hydrogen diffusion in the thermoplastic layers, simulating the movement of moisture in the annulus through the capillary effect, and incorporating fracture toughness degradation for a more thorough understanding of the interplay between mechanical and environmental factors in the hydrogen embrittlement of steel-reinforced RTP composite pipes.

Predicting corporate restructuring and financial distress in banks: The case of the Swiss banking industry

AbstractThe global financial crisis of 2007–2009 is widely regarded as the worst since the Great Depression and threatened the global financial system with a total collapse. This article distinguishes itself from the vast literature of bankruptcy, bank failure, and bank exit prediction models by introducing novel categorical parameters inspired by Switzerland's banking landscape. We evaluate data from 274 banks in Switzerland from 2007 to 2017 using generalized linear model logit and multinomial logit regressions and examine the determinants of corporate restructuring and financial distress. We complement our results with a robustness test via a Bayesian inference framework. We find that total assets and net interest margin affect bank exit and mergers and acquisitions, and that banks operating in the Zurich area have a higher likelihood of exiting and becoming takeover targets relative to banks operating in the Geneva area.

Energy harvester reliability study by Gaidai reliability method

AbstractThis study validates a novel structural reliability method, particularly suitable for high‐dimensional green energy harvesting device dynamic systems, versus a well‐established bivariate statistical method, known to accurately predict two‐dimensional system extreme response contours. Classic reliability methods dealing with time series do not always have an advantage of dealing easily with dynamic system high dimensionality, along with complex cross‐correlations among different system components. Energy harvesters constitute an important part of modern offshore green energy engineering; hence, proper experimental study along with safety and reliability analysis are of practical design and engineering importance. To study the performance of galloping energy harvesters, a series of laboratory wind tunnel tests have been conducted, selecting different wind speeds. This study illustrates the usage of the advocated novel reliability method, by analyzing bivariate statistics of experimental galloping energy harvester's dynamics. The bivariate statistics was extracted from available experimental results, more specifically for the device's voltage‐force dataset. Advantage of the proposed methodology being that relatively short experimental data record may still yield meaningful design results, provided proper statistical methods have been applied. Safety and reliability are important engineering concerns for all kinds of green energy devices. In the case of measured device's structural response, an accurate prediction of system failure or damage probability is possible, as illustrated in this study. Distinctive advantage of advocated novel semi‐analytical reliability methodology being the fact that it can tackle dynamic systems with practically unlimited number of dimensions (or components), along with complex nonlinear cross‐correlations between different system key components.

Conception of a new quality control method based on neural networks

The prediction of failures in a factory is now an important area of industry that helps to reduce time and cost of non-quality from the data generated from the sensors installed on production lines, this data is used to detect anomalies and predict defects before they occur. The purpose of this article is to model an intelligent production line capable of predicting various types of non-conforming products. For that, we will utilize the neural network methodology within the specific context of a production line specialized in juice manufacturing. Firstly, we introduce the production line under study, along with its distinct manufacturing phases. Secondly, we evaluate the performance indicators of this line, enabling us to gain an overview of its efficiency and overall performance. Subsequently, we present common industrial solutions that are frequently implemented to address the issues identified during our analysis. At this stage, we propose a predictive model based on neural network methodology. This model will possess the capability to detect and identify defective products and potential hazards within a production line before they occur. Throughout this study, we compare between three models of neural networks: LSTM model using Stochastic gradient descend (SGD), Feed forward model using ADAM Optimization and Feed forward model using Levenberg–Marquardt back propagation (LMBP), in order to determine the most optimal method in terms of achieved results. Finally, we demonstrate the effectiveness, performance, and accuracy of the results through the testing phase of the neural networks.

A Framework for Monitoring Stability of Tailings Dams in Realtime Using Digital Twin Simulation and Machine Learning

Tailings dam failures cause catastrophic impact on the environment and surrounding communities. Incidences of failure in the recent past have caused industrialists and researchers to seek innovative ways for proactively managing their safety and disaster mitigation. Given Industry 4.0 technologies now available, researchers are looking to develop digital tools for cost-effective, realtime monitoring of tailings dams. However, published literature indicates that a reliable framework is still lacking. This paper proposes a framework for developing a data-driven system for monitoring tailings dam stability and early warning detection. The framework relies upon digital twin simulation and machine-learning (ML) techniques, and comprises four main components: realtime data collection, digital twin modelling, ML-based early detection and prediction, and intelligence-driven decision-support. Sensors gather real-time geophysical data from monitored structure, and the digital twin uses this data to simulate dam behaviour. ML algorithms analyse the data and simulations to enable early detection of instability and failure prediction. Literature suggests that digital twin and ML-based approaches may have advantages over traditional monitoring techniques and other AI-based methods. The paper concludes with a discussion of the framework's limitations, opportunities for improvement, and potential for application in mining and geotechnical engineering. The paper serves as a basis for model development and future research.

Failure prediction in the refinery piping system using machine learning algorithms: classification and comparison

Pipelines play a pivotal role in transporting large volumes of oil and gas within refineries. However, over time, they are susceptible to deterioration, leading to potential failures. Effective monitoring is imperative to maintain their optimal performance and safety. This research introduces a machine learning (ML) approach to pinpoint failure sources in oil and gas pipelines. Analysing an industrial dataset, we compared six ML models to predict failures in refinery pipelines. Leakage sources are predicted based on three operational parameters: transported fluid, temperature, and pressure. The models are evaluated and compared in terms of precision, recall, F1-score, accuracy, and the ROC-AUC. Remarkably, the XGBoost classifier exhibited a 99.7% accuracy, outperforming other algorithms in predicting the failure source. Emphasizing the value of Industry 4.0 solutions, this study underscores the potential of advanced ML in enhancing pipeline monitoring. Such predictions empower operators to pre-empt failures, reinforcing industry safety and sustainability.

Precursor Prediction and Early Warning of Power MOSFET Failure Using Machine Learning with Model Uncertainty Considered

Precursor Prediction and Early Warning of Power MOSFET Failure Using Machine Learning with Model Uncertainty Considered

Utility of Machine Learning in the Prediction of Post-Hepatectomy Liver Failure in Liver Cancer.

Posthepatectomy liver failure (PHLF) is a serious complication associated with high mortality rates. Machine learning (ML) has rapidly developed and may outperform traditional models in predicting PHLF in patients who have undergone hepatectomy. This study aimed to predict PHLF using ML and compare its performance with that of traditional scoring systems. The clinicopathological data of 334 patients who underwent liver resection were retrospectively collected. The Pycaret library, a simple, open-source machine learning library, was used to compare multiple classification models for PHLF prediction. The predictive performance of 15 ML algorithms was compared using the mean area under the receiver operating characteristic curve (AUROC) and accuracy, and the best-fit model was selected among 15 ML algorithms. Next, the predictive performance of the selected ML-PHLF model was compared with that of routine scoring systems, the albumin-bilirubin score (ALBI) and the fibrosis-4 (FIB-4) index, using AUROC. The best model was extreme gradient boosting (accuracy:93.1%; AUROC:0.863) among the 15 ML algorithms. As compared with ALBI and FIB-4, the ML PHLF model had higher AUROC for predicting PHLF. The novel ML model for predicting PHLF outperformed routine scoring systems.

Enhancing the fire-resistance performance of composite laminates via multi-scale hybridisation: A review

Fibre-reinforced composites laminates (FRCLs) are employed in various applications such as in marine, aerospace, automotive, and civil industries due to their lightweight nature, design tailorability, and superior specific mechanical properties. However, they possess extremely low flame resistance mainly due to the inherent flammability of the polymer matrix. Various treatments have been applied to improve the fire resistance of FRCLs. In particular, hybridisation (fibre hybridisation and polymer hybridisation) is an important technique which is becoming very popular to enhance the thermal performance and flame resistance of FRCLs. This article is a comprehensive review of the recent developments that broadly cover the improvements in fire resistance of composite laminates via multi-scale hybridisation; the characteristics of thermal decomposition of FRCLs have been presented to comprehend the need for flame retardancy. Approaches for improving the fire resistance of FRCLs and thermal stability, both in polymer and in fibre systems, are discussed. Enhancing the fire resistance has been significant through additives to the matrix, use of flame-retardant modified fibres at interfacial regions and by way of multi-layered hybrid laminates besides hybridization at fibre, yarn and layer level. Finally, a review is presented on the modelling of fire resistance of composite laminates by considering thermo-mechanical models for the prediction of decomposition and failure of laminates at elevated temperatures.

Дослідження шляхів підвищення оперативності виконання діагностичних операцій електронних систем автотракторної техніки

The purpose of this study is to find ways to reduce production time spent on diagnostic operations during maintenance or repair of electronic control systems (ECS) in automotive vehicles. Existing methods of diagnosing engine control systems in vehicles are sufficiently effective, but in most cases, they require highly skilled personnel, the use of additional complex and expensive equipment, prolonged service cycles, which in turn increases the complexity of maintenance, and also does not provide for prior prediction of faults and system failures. Analysis of scientific publications has shown that to reduce the time for maintenance and repair of electronic systems, it is necessary to predict the residual life of ECS components in vehicles during maintenance, to further reduce the labor intensity of diagnostics and current repairs. In other words, diagnostics performed during maintenance or repair should include predicting the residual life of electronic system components. For the most critical ECS components in vehicles, it is proposed to apply a strategy of replacing them upon reaching a certain mileage, rather than upon failure. This organization method will reduce vehicle downtime in repair and reduce the time to troubleshoot ECS component failures. A method has been proposed for identifying design elements that limit the reliability of the ECS, taking into account the degree of influence of their failures on the engine's technical and economic characteristics and the cost of restoring its performance. The availability of such information is the basis of the ECS maintainability support system. An algorithm has been developed to search for hidden faults in the components of ECS subsystems, which includes predicting the failure of ECS components. This will help reduce the labor intensity of diagnostic operations during maintenance and repair.

Improving the reliability of power distribution substations equipment for life extension by considering Condition Based Maintenance

Maintenance of aged power equipment is a proven technique for establishing systems reliability and longevity. It is extensively relevant when combined with Condition Based Maintenance (CBM) approaches for predicting power system infrastructure life, especially with imminent obsolescence. The research focuses on distribution companies with mixed aged power infrastructure, which continuously degradate. A significant setback in older systems is the scarcity of replacement parts leading to unplanned investment in substantial infrastructure due to unexpected failed components that are no longer in production. To guarantee excellent infrastructure health in older equipment, utility companies should consider applying Condition Based Maintenance (CBM) using Artificial Neural Network (ANN) to stabilize older equipment and restore malfunctioning assets. Studies have shown that the life span of older equipment can be extended when adequate maintenance strategies are put in place through the prediction of possible failures using equipment inspections and risk matrix. CBM's strategy improved the Key Performance Indicator (KPIs) System Average Interruption Duration Index (SAIDI) has improved by lowering the average monthly customers power interruptions by 19,15 % lesser time spent without electricity. The System Average Interruption Frequency Index (SAIFI) was improved by lowering the average monthly power interruptions frequency by 20,63% lesser outage frequency. The Distribution Supply Loss Index (DSLI) was improved by dropping the average monthly power losses by 7,5% power losses saved. The Customers Average Interruption Duration Index (CAIDI) was improved by lowering the average customer interruption by 36.1% improvement from the average period before the CBM optimization. The Average System Availability Index (ASAI) was improved by shortening the average annual interruption durations by 1,36%. The ASUI was improved by increasing the average duration of electrical power availability by 1.28%. The optimization of CBM was measured over 24 moths period.

Prediction of retail chain failure: examples of recent U.S. retail failures

Prediction of retail chain failure: examples of recent U.S. retail failures

Hyper-Parameter Optimization through Reinforcement Learning for Survival Prediction of Patients with Heart Failure

Cardiovascular diseases (CVDs) are a class of medical conditions characterized by impaired functionality of the heart and blood vessels, among which coronary heart disease, cerebrovascular diseases, and heart failure are prominent. These diseases account for approximately one-third of all deaths recorded worldwide. A significant challenge currently faced by clinicians is the early prediction of heart failure. The employment of Machine Learning techniques has emerged as a leading method in predicting patient survival who exhibit symptoms of heart failure. Nonetheless, achieving a significant level of accuracy remains a topic that demands further investigation. The present study introduces a novel approach that leverages Reinforcement Learning (RL) to enhance the performance of the artificial neural network (ANN) techniques for survival prediction by identifying the optimal configuration of model hyper-parameters (HPO). The proposed approach has been assessed using a conventional benchmark dataset comprising the medical records of 299 patients. The results were compared with recent ANN approaches working on the same dataset. The findings indicate that the proposed approaches enhance the performance of the ANN-based heart failure predictive model.

Evaluation and Prediction of Post-Hepatectomy Liver Failure Using Imaging Techniques: Value of Gadoxetic Acid-Enhanced Magnetic Resonance Imaging.

Despite improvements in operative techniques and perioperative care, post-hepatectomy liver failure (PHLF) remains the most serious cause of morbidity and mortality after surgery, and several risk factors have been identified to predict PHLF. Although volumetric assessment using imaging contributes to surgical simulation by estimating the function of future liver remnants in predicting PHLF, liver function is assumed to be homogeneous throughout the liver. The combination of volumetric and functional analyses may be more useful for an accurate evaluation of liver function and prediction of PHLF than only volumetric analysis. Gadoxetic acid is a hepatocyte-specific magnetic resonance (MR) contrast agent that is taken up by hepatocytes via the OATP1 transporter after intravenous administration. Gadoxetic acid-enhanced MR imaging (MRI) offers information regarding both global and regional functions, leading to a more precise evaluation even in cases with heterogeneous liver function. Various indices, including signal intensity-based methods and MR relaxometry, have been proposed for the estimation of liver function and prediction of PHLF using gadoxetic acid-enhanced MRI. Recent developments in MR techniques, including high-resolution hepatobiliary phase images using deep learning image reconstruction and whole-liver T1 map acquisition, have enabled a more detailed and accurate estimation of liver function in gadoxetic acid-enhanced MRI.

СИСТЕМА УПРАВЛЕНИЯ БЕЗОПАСНОСТЬЮ ПОЛЁТОВ ЛЕТАТЕЛЬНЫХ АППАРАТОВ С БАЗОВОЙ ВСТРОЕННОЙ СИСТЕМОЙ АВТОМАТИЧЕСКОГО ДОПУСКОВОГО КОНТРОЛЯ ГИДРАВЛИЧЕСКОЙ СИСТЕМЫ И СИСТЕМ МОНИТОРИНГА И ПРОГНОЗИРОВАНИЯ ОТКАЗОВ

This paper considers an integrated tolerance control system with failure monitoring and prediction subsystems for aircraft safety management system (ASMS), which is relevant nowadays. The paper deals with the operational component of hydraulic system (HS) tolerance zones. The internal measured parameter is pressure, so it was necessary to find an algorithm that reflects its dependence on operating conditions, its coefficient of kinematic viscosity of the fluid, its temperature, operating time, ambient temperature and flight stage. The operational component of the tolerance zone can be derived from the obtained expression for the pressure by substituting the boundary values of the parameters included in this dependence. While polling the sensors of the automatic control system it is necessary each time to calculate the range of variation of the parameter for the given stage of flight and other conditions with the help of algorithms for calculating the tolerance zone obtained in the paper. Moreover, the control system has to process a large amount of information using artificial intelligence (AI) methods which allows the safety of aircraft flight (SAF) to be managed.

Performance Comparison of Decision Tree and Support Vector Machine Algorithms for Heart Failure Prediction

Heart failure is a significant global cause of mortality and morbidity. This research article aims to evaluate the performance of decision tree (DTree) and support vector machine (SVM) methods in predicting heart disease. The study utilizes a dataset with diverse features and employs exploratory data analysis (EDA), clustering, and classification techniques to gain insights and evaluate the performance of the two methods. The results demonstrate the effectiveness of both DTree and SVM in predicting heart disease. Notably, SVM outperforms DTree in terms of accuracy, precision, recall, and F1-score. However, the performance of these methods is influenced by the preprocessing steps applied, indicating the importance of selecting appropriate data preprocessing techniques for optimal performance with specific machine learning algorithms. This study emphasizes the potential of machine learning algorithms in predicting heart disease and underscores the significance of thoughtful preprocessing technique selection to enhance performance.

ОГЛЯД МЕХАНІЗМІВ ROOT CAUSE АНАЛІЗУ ТЕЛЕКОМУНІКАЦІЙНИХ МЕРЕЖ

Telecommunication networks are an integral part of modern information society, providing data transmission and communication among millions of users worldwide. The complexity and scale of these networks are constantly growing, which increases the demands on their reliability and stability. One of the key tasks in ensuring the efficient operation of telecommunication networks is the identification and elimination of the root causes of failures and anomalies that can significantly affect the quality of service for end users. Root cause analysis (RCA) is a powerful tool for identifying the primary causes of problems, preventing their recurrence, and enhancing the overall reliability of networks. The article provides a detailed review of contemporary methods and tools of RCA used in telecommunication networks. In particular, methods such as fishbone diagram analysis, the «five whys» method, fault tree analysis, and the application of machine learning for big data analysis are considered. Each of these approaches has its advantages and disadvantages, and their comprehensive use allows for higher accuracy and efficiency in identifying the root causes of problems. Practical examples demonstrate the effectiveness of applying various RCA methods to solve specific problems in telecommunication networks. For instance, the use of fishbone diagram analysis and the «five whys» method allows for the identification of the primary causes of network delays and equipment issues at base stations. The implementation of failure prediction systems based on machine learning significantly enhances network reliability by allowing for the early detection of potential problems and the necessary technical measures. The prospects for the development of RCA in telecommunication networks are outlined. The primary focus is on the development of adaptive and self-learning systems capable of dynamically responding to changing network conditions. The use of more complex artificial intelligence algorithms and tools for predicting potential problems will significantly improve the effectiveness of RCA and ensure high-quality telecommunication services.