{"title":"Addressing the label dilemma: A self-semi-supervised step-wise complementary label boosting strategy for industrial anomaly detection","authors":"Jiayang Yang, Chunhui Zhao","doi":"10.1016/j.ress.2025.111369","DOIUrl":"10.1016/j.ress.2025.111369","url":null,"abstract":"<div><div>Recently, Artificial Intelligence (AI) technology has been extensively employed in data-driven industrial anomaly detection. However, due to the difficulty of reliably acquiring the operating status of industrial processes, most process data may be collected without rigorous examination, resulting in uncertainty regarding their exact statuses and limiting their safe utilization for AI-powered anomaly detection modeling. Additionally, samples with definite annotations could still be subject to misjudgment of statuses by manual error, thereby exposing anomaly detection modeling to a significant risk of misleading. In this work, we accomplish anomaly detection as a binary classification task and recognize the aforementioned challenges as a modeling dilemma involving sample labels (annotations indicating their operating statuses, i.e., normal/abnormal), where the available labels are insufficient and unreliable simultaneously. Thereupon, a self-semi-supervised step-wise complementary label boosting (<span><math><msup><mrow><mi>S</mi></mrow><mrow><mn>4</mn></mrow></msup></math></span>CLB) strategy is proposed to address that dilemma. The <span><math><msup><mrow><mi>S</mi></mrow><mrow><mn>4</mn></mrow></msup></math></span>CLB strategy mainly consists of two stages, in the first stage, the self-supervised contrastive autoencoding Gaussian mixture model (CAGMM) is developed to provide representations of all the process samples for the subsequent anomaly detection by describing their data distribution information with low-dimensional features. In the second stage, a semi-supervised label boosting strategy is designed in a step-wise manner. Specifically, the noisy label filtering and adaptive label enrichment are conducted alternately to boost the sufficiency and reliability of available labels regressively. Meanwhile, the robust dual complementary classifier (RDCC) model comprising two peer classifiers with robustness and different views is developed to achieve the prompt feedback for label boosting, thus the reliability of label adjustment is further guaranteed. Finally, the anomaly detection results are obtained by the RDCC model. The effectiveness of the proposed method is verified by a real industrial process.</div></div>","PeriodicalId":54500,"journal":{"name":"Reliability Engineering & System Safety","volume":"264 ","pages":"Article 111369"},"PeriodicalIF":9.4,"publicationDate":"2025-06-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144510995","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Reliability prediction using a weighted temporal convolutional autoencoder based on limited claim data","authors":"Seong-Mok Kim , Min Jung , Yong Soo Kim","doi":"10.1016/j.ress.2025.111374","DOIUrl":"10.1016/j.ress.2025.111374","url":null,"abstract":"<div><div>Many product manufacturing companies offer warranty services that cover costs incurred due to product failures during the warranty period. To maximize savings on warranty costs, researchers have attempted to predict field reliability using short-term claim data. Due to the limited information available, however, the prediction performance for long warranty periods has been unsatisfactory. This study proposes a weighted temporal convolutional autoencoder (WTCAE) model designed to predict the number of claims and field reliability over the entire warranty period using limited initial claim data. The WTCAE model compensates for the limited information from initial claim data by effectively capturing temporal patterns through a temporal convolutional network-based encoder–decoder structure. The proposed WTCAE model demonstrated superior performance even under conditions of short-term claim data, where traditional lifetime distribution-based methods fail to provide predictions. It also consistently outperformed conventional deep learning-based methods. The effectiveness and practicality of the proposed WTCAE model were validated using real-world data from millions of televisions and refrigerators, confirming its consistent performance across various data conditions within the warranty period.</div></div>","PeriodicalId":54500,"journal":{"name":"Reliability Engineering & System Safety","volume":"264 ","pages":"Article 111374"},"PeriodicalIF":9.4,"publicationDate":"2025-06-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144511023","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Liye Zhang , Kewang Gu , Zhicheng Ma , Bing Wu , Jie Song
{"title":"Modelling collision risk between container and fishing ships during cross encounter in a cordon area","authors":"Liye Zhang , Kewang Gu , Zhicheng Ma , Bing Wu , Jie Song","doi":"10.1016/j.ress.2025.111373","DOIUrl":"10.1016/j.ress.2025.111373","url":null,"abstract":"<div><div>Collisions between container ships and fishing ships frequently result in significant consequences, particularly in China. To assess the collision risk of ship cross-meetings in the warning area, this study constructed a collision risk model using historical AIS data. The risk influencing factors included relative distance, relative speed, relative bearing, distance closest point of approach (DCPA) and time closest point of approach (TCPA). We extracted pairs of trajectories of ships of different types (container ships and fishing ships) and employed the entropy approach to calculate the weights of each risk factor for the modelling of collisions between two ships in real time. Our findings indicate that the collision risk during a two-ship cross-meeting increases gradually as the meeting progresses, reaching a maximum value (approximately 0.66) at the point of nearest proximity, after which it decreases. A comparison and analysis of the three areas revealed that the warning area represented a high-risk area for the cross-meeting. In order to calculate the weight value of each ship and assess the real-time collision risk during the multi-ship cross-meeting, the Shapley value method was employed. The results indicate that the collision risk of a multi-ship encounter is significantly higher than that of a two-ship encounter. Furthermore, the change in risk values is more complex. This study not only quantifies the risk between two ships, but also provides a new method to quantify the risk of multiple ships. This provides important theoretical support for the safety of ship navigation.</div></div>","PeriodicalId":54500,"journal":{"name":"Reliability Engineering & System Safety","volume":"264 ","pages":"Article 111373"},"PeriodicalIF":9.4,"publicationDate":"2025-06-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144517762","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Yongchao Zhang , Zhiyuan Wang , Caizi Fan , Zeyu Jiang , Kun Yu , Zhaohui Ren , Ke Feng
{"title":"Diffusion model-assisted cross-domain fault diagnosis for rotating machinery under limited data","authors":"Yongchao Zhang , Zhiyuan Wang , Caizi Fan , Zeyu Jiang , Kun Yu , Zhaohui Ren , Ke Feng","doi":"10.1016/j.ress.2025.111372","DOIUrl":"10.1016/j.ress.2025.111372","url":null,"abstract":"<div><div>In industrial scenarios, cross-domain fault diagnosis faces the challenge of data scarcity due to the difficulty of data acquisition and the high cost of labeling. To overcome this issue, this paper proposes a diffusion model-assisted data generation method to enhance the model’s cross-domain diagnostic capability by generating target domain data. Specifically, this paper establishes a diffusion model-assisted cross-domain fault diagnosis method, where a diffusion model is first constructed to augment the target domain data, and then a deep learning model is jointly trained using source domain data, a small amount of target domain data, and the generated target domain data to learn and transfer diagnostic knowledge. To align the global feature distributions, the maximum mean discrepancy loss is first employed to align the source domain data with both the target domain data and the generated target domain data. Additionally, a cross-domain triplet loss is established to achieve category alignment and separation, ensuring similar categories are aligned while different categories are distinguished. Finally, the deep consistency regularization is designed to enforce consistency across target domain data and its augmented versions, enhancing the model’s robustness. Extensive experiments on two rotating machinery systems demonstrate the effectiveness of the proposed method in addressing limited-data cross-domain fault diagnosis, highlighting its potential for practical applications in intelligent health monitoring of rotating machinery.</div></div>","PeriodicalId":54500,"journal":{"name":"Reliability Engineering & System Safety","volume":"264 ","pages":"Article 111372"},"PeriodicalIF":9.4,"publicationDate":"2025-06-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144517763","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Joint optimization of condition-based maintenance and component reallocation for phased-mission balanced systems with flexible structure considering imperfect substitution","authors":"Siqi Wang , Shuyun Li , Lipo Mo","doi":"10.1016/j.ress.2025.111399","DOIUrl":"10.1016/j.ress.2025.111399","url":null,"abstract":"<div><div>In many practical situations, many balanced systems need to operate in multiple phases. Such systems usually have a different system structure for each phase. The system is balanced when the maximum state difference between all working components is less than a predetermined value. Component degradation is described by a Markov process and the state can be checked periodically. We propose a joint policy of condition-based component reallocation and maintenance for this balanced system. Non-working components can be used to replace working components for operation. Components can be repaired to an arbitrarily better state. The operation process of the balanced system is described as a Markov decision process and the optimal joint policy is obtained by a finite stage backward recursive iterative algorithm. In addition, two contrasting policies are proposed to compare with the proposed policy. Based on flexible manufacturing systems, a numerical study shows that the proposed joint policy is significantly better than the other two policies.</div></div>","PeriodicalId":54500,"journal":{"name":"Reliability Engineering & System Safety","volume":"264 ","pages":"Article 111399"},"PeriodicalIF":9.4,"publicationDate":"2025-06-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144517761","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Chang Wang , Hua Zhou , Sen Lin , Xiaodan Weng , Yu Shao , Tingchao Yu
{"title":"A hybrid deep survival model for failure modeling of water distribution networks coupling physical survival and data reconstruction","authors":"Chang Wang , Hua Zhou , Sen Lin , Xiaodan Weng , Yu Shao , Tingchao Yu","doi":"10.1016/j.ress.2025.111401","DOIUrl":"10.1016/j.ress.2025.111401","url":null,"abstract":"<div><div>To detect and replace damaged pipes and maintain the stable operation of water supply systems timely, it is crucial to carry out pipeline failure prediction. Limited by the feature nonlinear ability and generalization ability of survival machine learning, the application effect in pipeline failure prediction is not always satisfactory. To develop more efficient, powerful, and flexible prediction models, a hybrid deep survival model (HDSM) is proposed by coupling deep auto-encoder and survival analysis to effectively predict pipeline failures. Guided by the theory of mechanism and data fusion, the data reconstruction constraints and survival analysis constraints are coupled into the loss function by HDSM. Its dual advantages of combining the powerful feature extraction capability of deep auto-encoder and the statistical inference role of survival analysis can more accurately predict the survival time of physical pipeline systems. In a real pipeline network, the superiority and effectiveness of HDSM are verified in comparison with other deep survival learning and survival machine learning, with C-index exceeding 0.95 and Brier score below 0.056. Finally, sensitivity analyses of different hyperparameters are carried out to verify the robustness of the HDSM model in pipeline failure prediction.</div></div>","PeriodicalId":54500,"journal":{"name":"Reliability Engineering & System Safety","volume":"264 ","pages":"Article 111401"},"PeriodicalIF":9.4,"publicationDate":"2025-06-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144517760","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Dong Xu , Yubin Tian , Dianpeng Wang , Junbiao Shi
{"title":"Reliability analysis and layout optimization for a multi-component system with thermal coupling","authors":"Dong Xu , Yubin Tian , Dianpeng Wang , Junbiao Shi","doi":"10.1016/j.ress.2025.111348","DOIUrl":"10.1016/j.ress.2025.111348","url":null,"abstract":"<div><div>An important feature of power and electronic devices is that their operation is accompanied by the release of heat, which leads to thermal coupling between components, that is, the interaction of temperatures between adjacent components. This phenomenon reflects spatial dependence and is rarely considered in reliability analyses. In this study, a reliability model was proposed for a multi-component system with thermal coupling and was subsequently extended to a competing failure model. Additionally, considering that different components have different workloads, components with higher workloads should be located further away from each other to reduce the probability of high temperatures caused by the simultaneous operation of the components, thus increasing the system’s reliability. Through the innovative use of the minimum energy criterion, we present a layout optimization approach to this issue. Furthermore, the larger the component spacing, the weaker the thermal coupling effect, the higher the system reliability, and the bulkier the system. Therefore, a trade-off must be made. A redundancy allocation problem was studied, that is, minimizing the system volume while considering a given reliability constraint. A numerical example demonstrates the effectiveness of layout optimization in improving reliability and illustrates the application of the proposed methods.</div></div>","PeriodicalId":54500,"journal":{"name":"Reliability Engineering & System Safety","volume":"264 ","pages":"Article 111348"},"PeriodicalIF":9.4,"publicationDate":"2025-06-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144510984","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Topological analysis of risks in hazardous materials transportation systems using fitness landscape theory and association rules mining","authors":"Jian Guo , Kaijiang Ma , Haoxuan Ren","doi":"10.1016/j.ress.2025.111396","DOIUrl":"10.1016/j.ress.2025.111396","url":null,"abstract":"<div><div>Determining the failure modes of hazardous materials transportation systems, considering the coupled effects of risk factors, is crucial for ensuring transportation safety. This study proposes a coupled topological analysis method for hazardous materials road transport risks, based on association rule mining and fitness landscape theory. This method can reflect the correlations and evolutionary patterns of risk factors, thereby providing a basis for formulating risk mitigation strategies. Firstly, text mining techniques are employed to identify critical risk factors and gather a structured dataset comprising 165 entries. Secondly, association rule algorithms are used to uncover potential relationships among sub-factors, employing the Apriori algorithm with set thresholds to extract strong association rules, which are then mapped into a landscape model depicting the coupled evolution of system risk factors. Finally, by employing a defined fitness function, typical system failure paths are further analyzed topologically. The results indicate that directly mining failure paths from sub-risk factors can elucidate more detailed system failure mechanisms. Coupled failure modes involving human and environmental factors warrant particular attention. Vehicle factors often lead to accidents without further evolution, necessitating the establishment of corresponding inspection mechanisms.</div></div>","PeriodicalId":54500,"journal":{"name":"Reliability Engineering & System Safety","volume":"264 ","pages":"Article 111396"},"PeriodicalIF":9.4,"publicationDate":"2025-06-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144510985","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Advanced statistical analysis and system reliability assessment of API 5L steel pipelines subjected to corrosion attack","authors":"Djamel Zelmati , Omar Bouledroua , Oualid Ghelloudj , Riad Harouz","doi":"10.1016/j.ress.2025.111387","DOIUrl":"10.1016/j.ress.2025.111387","url":null,"abstract":"<div><div>It is crucial to note that developing a maintenance and repair strategy by statistically consolidating the corrosion features across the pipeline length ultimately yields a single probability of failure for the entire line, potentially obscuring the distinct statistical behavior of individual pipe segments. Since the corrosion profiles in different segments of the pipeline are statistically correlated, failure in one section can affect neighboring sections, which defines the overall system reliability. In this study, the remaining strength of a corroded pipeline was estimated using the Failure Assessment Diagram (FAD). A comprehensive statistical analysis identified the typical probability density functions of all random variables in the failure scenario, as well as their standard deviations and typical correlation coefficients. A Monte Carlo simulation (MCS) was integrated with the SINTAP procedure to construct a probabilistic FAD. Sensitivity analysis revealed the relative influence of each variable on the pipeline’s remaining strength, emphasizing a strong interaction between defect depth and wall thickness. Additionally, the combined effect of the operating pressure and the coefficient of variation of the corrosion defect depth was evaluated to understand their influence on the probability of failure. Furthermore, system reliability assessment allows for more precise risk management, avoiding overly conservative or optimistic estimates that might arise from treating the entire pipeline as a single unit. Besides, the developed procedure for global system reliability can be used to devise optimal and more accurate inspection and maintenance schedules.</div></div>","PeriodicalId":54500,"journal":{"name":"Reliability Engineering & System Safety","volume":"264 ","pages":"Article 111387"},"PeriodicalIF":9.4,"publicationDate":"2025-06-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144489381","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Reliability assessment methods considering failure correlation and importance analysis for multiple progressive damage","authors":"Zhixuan Gao , Deyin Jiang , Xinchen Zhuang , Weimin Cui , Tianxiang Yu","doi":"10.1016/j.ress.2025.111389","DOIUrl":"10.1016/j.ress.2025.111389","url":null,"abstract":"<div><div>Aircraft complex mechanisms are susceptible to multiple progressive damage in operation, where a particular form of damage can have an effect on the strength of another through the mechanism transmission. The results of reliability analysis that ignores the correlation between them will deviate from the actual situation. To address the above problems, based on the characteristics of multiple progressive damage, two correlation models of multiple progressive damage are established, and a reliability assessment method considering the correlation of multiple progressive damage is proposed. Based on the vine copula theory, the correlation model of multiple progressive damage systems at different times was established, and the reliability of the system was calculated based on Monte Carlo method. Based on the importance theory, the importance analysis method is further proposed for multiple progressive damage. The results of the engineering case study show that ignoring the correlation between multiple progressive damage can lead to an inaccurate system reliability assessment, the maximum relative error to the reliability results for the independent case is 25 %. And the types of damage that need to be focused on maintenance are obtained by importance analysis, and the method provides a reference for the reliability assessment of complex mechanisms in aerospace.</div></div>","PeriodicalId":54500,"journal":{"name":"Reliability Engineering & System Safety","volume":"264 ","pages":"Article 111389"},"PeriodicalIF":9.4,"publicationDate":"2025-06-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144510986","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}