Advanced Engineering Informatics最新文献_第2页

CPIR: Multimodal Industrial Anomaly Detection via Latent Bridged Cross-modal Prediction and Intra-modal Reconstruction

IF 8 1区工程技术 Q1 COMPUTER SCIENCE, ARTIFICIAL INTELLIGENCE

Advanced Engineering Informatics

Pub Date : 2025-03-14 DOI: 10.1016/j.aei.2025.103240

Wen Shangguan , Hongqiang Wu , Yanchang Niu , Haonan Yin , Jiawei Yu , Bokui Chen , Biqing Huang

While RGB-based methods have been extensively studied in Industrial Anomaly Detection (IAD), effectively incorporating point cloud data remains challenging. Alongside prevalent memory bank-based approaches, recent research has explored cross-modal feature mapping for multimodal IAD, achieving notable performance and efficient inference. However, cross-modal feature mapping, while effective for detecting anomalies in feature correspondences, struggles to identify those exclusive to a single modality, due to the inherent one-to-many mapping between 2D and 3D data. To overcome this limitation, we propose Cross-modal Prediction and Intra-modal Reconstruction (CPIR), a novel multimodal anomaly detection method. First, we introduce a Bidirectional Feature Mapping (BFM) framework that integrates intra-modal reconstruction tasks with cross-modal prediction tasks, enhancing single-modality anomaly detection while maintaining effective cross-modal consistency learning. Second, we propose a novel network architecture, Latent Bridged Modal Mapping Module (LB3M), which introduces a shared latent intermediate state to decouple feature mapping across modalities into mappings between each modality and a shared intermediate state. This design was initially proposed to effectively complete prediction and reconstruction tasks with minimal parameters. However, it also enabled the network to learn more comprehensive feature patterns, significantly improving anomaly detection capabilities. Experiments on the MVTec 3D-AD dataset demonstrate that CPIR outperforms state-of-the-art methods in both anomaly detection and segmentation tasks, while excelling in few-shot learning scenarios.

{"title":"CPIR: Multimodal Industrial Anomaly Detection via Latent Bridged Cross-modal Prediction and Intra-modal Reconstruction","authors":"Wen Shangguan , Hongqiang Wu , Yanchang Niu , Haonan Yin , Jiawei Yu , Bokui Chen , Biqing Huang","doi":"10.1016/j.aei.2025.103240","DOIUrl":"10.1016/j.aei.2025.103240","url":null,"abstract":"<div><div>While RGB-based methods have been extensively studied in Industrial Anomaly Detection (IAD), effectively incorporating point cloud data remains challenging. Alongside prevalent memory bank-based approaches, recent research has explored cross-modal feature mapping for multimodal IAD, achieving notable performance and efficient inference. However, cross-modal feature mapping, while effective for detecting anomalies in feature correspondences, struggles to identify those exclusive to a single modality, due to the inherent one-to-many mapping between 2D and 3D data. To overcome this limitation, we propose <strong>Cross-modal Prediction and Intra-modal Reconstruction (CPIR)</strong>, a novel multimodal anomaly detection method. First, we introduce a <strong>Bidirectional Feature Mapping (BFM)</strong> framework that integrates intra-modal reconstruction tasks with cross-modal prediction tasks, enhancing single-modality anomaly detection while maintaining effective cross-modal consistency learning. Second, we propose a novel network architecture, <strong>Latent Bridged Modal Mapping Module (LB3M)</strong>, which introduces a shared latent intermediate state to decouple feature mapping across modalities into mappings between each modality and a shared intermediate state. This design was initially proposed to effectively complete prediction and reconstruction tasks with minimal parameters. However, it also enabled the network to learn more comprehensive feature patterns, significantly improving anomaly detection capabilities. Experiments on the MVTec 3D-AD dataset demonstrate that CPIR outperforms state-of-the-art methods in both anomaly detection and segmentation tasks, while excelling in few-shot learning scenarios.</div></div>","PeriodicalId":50941,"journal":{"name":"Advanced Engineering Informatics","volume":"65 ","pages":"Article 103240"},"PeriodicalIF":8.0,"publicationDate":"2025-03-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143619134","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}

引用次数: 0

A human-centric model for task demand assessment based on unsupervised learning-assisted eye movement measure

IF 8 1区工程技术 Q1 COMPUTER SCIENCE, ARTIFICIAL INTELLIGENCE

Advanced Engineering Informatics

Pub Date : 2025-03-14 DOI: 10.1016/j.aei.2025.103259

Bufan Liu , Sun Woh Lye , Kai Xiang Yeo , Chun-Hsien Chen

Nowadays, Industry 5.0 marks a transformative shift from the focus on efficiency to a human-centered approach, emphasizing the principles of human-AI hybrid systems. This mode prioritizes intelligent technology that supports human abilities rather than replacing them, especially in safety–critical fields like air traffic management (ATM) with human controllers playing an essential role in maintaining safe and efficient operations. Quantifying the task demand of air traffic controllers (ATCOs) is vital to ensure optimal taskload management, thereby assisting in mitigating the risk of human error and promoting sustained operational performance. To realize this aim, this research proposes a human-centric model for task demand assessment based on unsupervised learning-assisted eye movement measure. Initially, two data streams are gathered from human-in-the-loop visual tasks, capturing flight information on the radar screen and eye-tracking data. These data streams are then synchronized and merged by aligning their timestamps. Subsequently, an unsupervised learning-based clustering approach is implemented, utilizing the OPTICS model to identify areas of interest based on aircraft positions, along with the K-Means model to categorize task intensity levels using the derived eye movement data. Finally, a task demand score index is developed for each task intensity level and across all task categories, with parameter weights determined through an entropy-based method. Comprehensive results and analyses are presented to illustrate the method’s applicability and effectiveness. This research paves the way for quantitatively understanding the specific taskload placed on ATCOs.

{"title":"A human-centric model for task demand assessment based on unsupervised learning-assisted eye movement measure","authors":"Bufan Liu , Sun Woh Lye , Kai Xiang Yeo , Chun-Hsien Chen","doi":"10.1016/j.aei.2025.103259","DOIUrl":"10.1016/j.aei.2025.103259","url":null,"abstract":"<div><div>Nowadays, Industry 5.0 marks a transformative shift from the focus on efficiency to a human-centered approach, emphasizing the principles of human-AI hybrid systems. This mode prioritizes intelligent technology that supports human abilities rather than replacing them, especially in safety–critical fields like air traffic management (ATM) with human controllers playing an essential role in maintaining safe and efficient operations. Quantifying the task demand of air traffic controllers (ATCOs) is vital to ensure optimal taskload management, thereby assisting in mitigating the risk of human error and promoting sustained operational performance. To realize this aim, this research proposes a human-centric model for task demand assessment based on unsupervised learning-assisted eye movement measure. Initially, two data streams are gathered from human-in-the-loop visual tasks, capturing flight information on the radar screen and eye-tracking data. These data streams are then synchronized and merged by aligning their timestamps. Subsequently, an unsupervised learning-based clustering approach is implemented, utilizing the OPTICS model to identify areas of interest based on aircraft positions, along with the K-Means model to categorize task intensity levels using the derived eye movement data. Finally, a task demand score index is developed for each task intensity level and across all task categories, with parameter weights determined through an entropy-based method. Comprehensive results and analyses are presented to illustrate the method’s applicability and effectiveness. This research paves the way for quantitatively understanding the specific taskload placed on ATCOs.</div></div>","PeriodicalId":50941,"journal":{"name":"Advanced Engineering Informatics","volume":"65 ","pages":"Article 103259"},"PeriodicalIF":8.0,"publicationDate":"2025-03-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143627980","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}

引用次数: 0

Multi-source ensemble transfer learning-based unmanned aerial vehicle flight data anomaly detection with limited data: From simulation to reality

IF 8 1区工程技术 Q1 COMPUTER SCIENCE, ARTIFICIAL INTELLIGENCE

Advanced Engineering Informatics

Pub Date : 2025-03-14 DOI: 10.1016/j.aei.2025.103255

Lei Yang , Shaobo Li , Caichao Zhu , Jian Liu , Ansi Zhang

Flight data anomaly detection is critical for ensuring the safety and reliability of unmanned aerial vehicles (UAVs). Traditional deep learning methods excel when sufficient data is available, but their performance significantly diminishes in data-scarce scenarios. Transfer learning is a promising solution; however, the performance of single-source transfer methods is often limited when there is a significant discrepancy between the source and target domains. This paper proposes a multi-source ensemble transfer learning-based anomaly detection (MSETL-AD) framework, aiming to transfer knowledge from multiple simulated domains to a real domain for anomaly detection in UAV flight data with limited data. First, a similarity calculation method based on dynamic time warping (DTW) is utilized to select simulated source domains that are similar to the target domain to mitigate the negative transfer problem. Second, a modeling strategy based on long short-term memory with attention mechanism (LSTM-AM) integrating transfer learning and fine-tuning techniques is proposed, which constructs a fundamental LSTM-AM prediction model for each source domain and then fine-tunes it using limited data in the target domain during the transfer process. Then, a similarity-based transfer weight assignment method is designed to guide multi-source domains for integration. Next, a similarity-guided dynamic threshold calculation method based on extreme value theory with residual smoothing is introduced to overcome random noise interference and realize adaptive anomaly detection. Finally, the effectiveness of the proposed method is validated through experiments using multiple simulated UAV flight datasets as the source domains and a real UAV flight dataset as the target domain.

{"title":"Multi-source ensemble transfer learning-based unmanned aerial vehicle flight data anomaly detection with limited data: From simulation to reality","authors":"Lei Yang , Shaobo Li , Caichao Zhu , Jian Liu , Ansi Zhang","doi":"10.1016/j.aei.2025.103255","DOIUrl":"10.1016/j.aei.2025.103255","url":null,"abstract":"<div><div>Flight data anomaly detection is critical for ensuring the safety and reliability of unmanned aerial vehicles (UAVs). Traditional deep learning methods excel when sufficient data is available, but their performance significantly diminishes in data-scarce scenarios. Transfer learning is a promising solution; however, the performance of single-source transfer methods is often limited when there is a significant discrepancy between the source and target domains. This paper proposes a multi-source ensemble transfer learning-based anomaly detection (MSETL-AD) framework, aiming to transfer knowledge from multiple simulated domains to a real domain for anomaly detection in UAV flight data with limited data. First, a similarity calculation method based on dynamic time warping (DTW) is utilized to select simulated source domains that are similar to the target domain to mitigate the negative transfer problem. Second, a modeling strategy based on long short-term memory with attention mechanism (LSTM-AM) integrating transfer learning and fine-tuning techniques is proposed, which constructs a fundamental LSTM-AM prediction model for each source domain and then fine-tunes it using limited data in the target domain during the transfer process. Then, a similarity-based transfer weight assignment method is designed to guide multi-source domains for integration. Next, a similarity-guided dynamic threshold calculation method based on extreme value theory with residual smoothing is introduced to overcome random noise interference and realize adaptive anomaly detection. Finally, the effectiveness of the proposed method is validated through experiments using multiple simulated UAV flight datasets as the source domains and a real UAV flight dataset as the target domain.</div></div>","PeriodicalId":50941,"journal":{"name":"Advanced Engineering Informatics","volume":"65 ","pages":"Article 103255"},"PeriodicalIF":8.0,"publicationDate":"2025-03-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143619132","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}

引用次数: 0

Infer potential accidents from hazard reports: A causal hierarchical multi label classification approach

IF 8 1区工程技术 Q1 COMPUTER SCIENCE, ARTIFICIAL INTELLIGENCE

Advanced Engineering Informatics

Pub Date : 2025-03-13 DOI: 10.1016/j.aei.2025.103237

Yipu Qin, Xinbo Ai

Inferring categories of accidents caused by hidden hazards detected contributes to safety management and prevention of accidents. For the safety management of many enterprises in a large administrative area, it is necessary to rely on industry experts to review hazard reports produced by front-line employees and infer the categories of potential accidents, which is time-consuming and labor-intensive. In this study, a hierarchical multi-label classification model is proposed to learn a checklist reviewed by industry experts and realize the automatic inference of the accident categories based on hazard descriptions. We simultaneously use the causal effect estimation method designed according to the backdoor adjustment in causal theory to extract the causal part of the text that affects the inference and design a data augmentation method based on the discovered causal knowledge to make the model focus on the causal key information to improve the inference and generalization abilities of the models. From the perspective of theoretical and practical contributions, this study not only realizes the estimation of causal effect of hazard words and the automatic inference of accident categories, which provides support for further accident prevention and safety management. It also makes a successful attempt to apply causality theory combined with deep learning methods in the field of safety, providing a valuable reference for future research on the combination of causal theory and practical applications.

{"title":"Infer potential accidents from hazard reports: A causal hierarchical multi label classification approach","authors":"Yipu Qin, Xinbo Ai","doi":"10.1016/j.aei.2025.103237","DOIUrl":"10.1016/j.aei.2025.103237","url":null,"abstract":"<div><div>Inferring categories of accidents caused by hidden hazards detected contributes to safety management and prevention of accidents. For the safety management of many enterprises in a large administrative area, it is necessary to rely on industry experts to review hazard reports produced by front-line employees and infer the categories of potential accidents, which is time-consuming and labor-intensive. In this study, a hierarchical multi-label classification model is proposed to learn a checklist reviewed by industry experts and realize the automatic inference of the accident categories based on hazard descriptions. We simultaneously use the causal effect estimation method designed according to the backdoor adjustment in causal theory to extract the causal part of the text that affects the inference and design a data augmentation method based on the discovered causal knowledge to make the model focus on the causal key information to improve the inference and generalization abilities of the models. From the perspective of theoretical and practical contributions, this study not only realizes the estimation of causal effect of hazard words and the automatic inference of accident categories, which provides support for further accident prevention and safety management. It also makes a successful attempt to apply causality theory combined with deep learning methods in the field of safety, providing a valuable reference for future research on the combination of causal theory and practical applications.</div></div>","PeriodicalId":50941,"journal":{"name":"Advanced Engineering Informatics","volume":"65 ","pages":"Article 103237"},"PeriodicalIF":8.0,"publicationDate":"2025-03-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143619713","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}

引用次数: 0

Evolving process maintenance through human-robot collaboration: An agent-based system performance analysis

IF 8 1区工程技术 Q1 COMPUTER SCIENCE, ARTIFICIAL INTELLIGENCE

Advanced Engineering Informatics

Pub Date : 2025-03-13 DOI: 10.1016/j.aei.2025.103241

Shuo yang , Micaela Demichela , Zhangwei Ling , Jie Geng

Periodic inspections of pressurized vessel systems are essential for maintaining safety through early fault detection. Traditional inspections often expose human operators to hazardous conditions within confined spaces. The advent of inspection robots has shifted the paradigm towards human-robot collaboration (HRC), which seeks to reduce risk while maintaining operational adaptability. This study compared the HRC and fully manual (FM) inspection processes, providing strategic insights for stakeholders. Historically, system performance evaluations have simplified or ignored dynamic human factors. To address this oversight, our research employs Agent-Based Models (ABMs) that encompass the evolving nature of human error, including the impact of fatigue and organizational factors, as well as the variability of human behavior and error recovery mechanisms. Our findings reveal that HRC significantly outperforms FM inspections, enhancing efficiency, accuracy, and safety. Notably, the study confirms that the miss rate of artificial intelligence (AI) for image identification within the HRC process is crucial for reliability and should not fall below the threshold of 0.04. This threshold is a benchmark for AI performance in HRC systems, ensuring that the balance between automated efficiency and human oversight is optimized. The research provides a comprehensive evaluation of HRC in pressurized vessel inspections. It offers a deeper understanding of the complex dynamics involved, advocating for integrating robust AI algorithms to support human operators in safety–critical tasks.

{"title":"Evolving process maintenance through human-robot collaboration: An agent-based system performance analysis","authors":"Shuo yang , Micaela Demichela , Zhangwei Ling , Jie Geng","doi":"10.1016/j.aei.2025.103241","DOIUrl":"10.1016/j.aei.2025.103241","url":null,"abstract":"<div><div>Periodic inspections of pressurized vessel systems are essential for maintaining safety through early fault detection. Traditional inspections often expose human operators to hazardous conditions within confined spaces. The advent of inspection robots has shifted the paradigm towards human-robot collaboration (HRC), which seeks to reduce risk while maintaining operational adaptability. This study compared the HRC and fully manual (FM) inspection processes, providing strategic insights for stakeholders. Historically, system performance evaluations have simplified or ignored dynamic human factors. To address this oversight, our research employs Agent-Based Models (ABMs) that encompass the evolving nature of human error, including the impact of fatigue and organizational factors, as well as the variability of human behavior and error recovery mechanisms. Our findings reveal that HRC significantly outperforms FM inspections, enhancing efficiency, accuracy, and safety. Notably, the study confirms that the miss rate of artificial intelligence (AI) for image identification within the HRC process is crucial for reliability and should not fall below the threshold of 0.04. This threshold is a benchmark for AI performance in HRC systems, ensuring that the balance between automated efficiency and human oversight is optimized. The research provides a comprehensive evaluation of HRC in pressurized vessel inspections. It offers a deeper understanding of the complex dynamics involved, advocating for integrating robust AI algorithms to support human operators in safety–critical tasks.</div></div>","PeriodicalId":50941,"journal":{"name":"Advanced Engineering Informatics","volume":"65 ","pages":"Article 103241"},"PeriodicalIF":8.0,"publicationDate":"2025-03-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143610255","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

An integrated approach for automatic safety inspection in construction: Domain knowledge with multimodal large language model

IF 8 1区工程技术 Q1 COMPUTER SCIENCE, ARTIFICIAL INTELLIGENCE

Advanced Engineering Informatics

Pub Date : 2025-03-13 DOI: 10.1016/j.aei.2025.103246

Yiheng Wang, Hanbin Luo, Weili Fang

This research addresses the challenge of dynamically integrating visual and textual data in construction safety inspections while enhancing adaptability to new safety hazards and ensuring faithful interpretation of safety rules. We propose a novel approach that seamlessly combines multi-modal techniques with domain knowledge, advancing beyond current methods that often struggle with multi-modal understanding and adaptation to new safety hazards. Our approach consists of three key components: (1) a fine-tuned multi-modal LLM for visual and textual processing, (2) a domain knowledge base for evolving safety standards adaptability and output faithfulness, and (3) a multi-step reasoning engine to tackle complex safety inspection tasks. We validate our approach using on-site data from Wuhan subway construction sites, demonstrating its capability to perform moderately accurate (0.57 hazard identification correctness), contextually relevant (0.96 on task relevancy), and faithful safety assessments (0.95 and 0.99 on reasoning faithfulness). The results suggest promising performance in construction scene perception, as well as textual analysis and reasoning. This approach represents an advancement in automatic construction safety inspection and contributes to the broader discourse on formalizing multi-modal processing of construction data, offering insights into creating more flexible and comprehensive safety management systems.

{"title":"An integrated approach for automatic safety inspection in construction: Domain knowledge with multimodal large language model","authors":"Yiheng Wang, Hanbin Luo, Weili Fang","doi":"10.1016/j.aei.2025.103246","DOIUrl":"10.1016/j.aei.2025.103246","url":null,"abstract":"<div><div>This research addresses the challenge of dynamically integrating visual and textual data in construction safety inspections while enhancing adaptability to new safety hazards and ensuring faithful interpretation of safety rules. We propose a novel approach that seamlessly combines multi-modal techniques with domain knowledge, advancing beyond current methods that often struggle with multi-modal understanding and adaptation to new safety hazards. Our approach consists of three key components: (1) a fine-tuned multi-modal LLM for visual and textual processing, (2) a domain knowledge base for evolving safety standards adaptability and output faithfulness, and (3) a multi-step reasoning engine to tackle complex safety inspection tasks. We validate our approach using on-site data from Wuhan subway construction sites, demonstrating its capability to perform moderately accurate (0.57 hazard identification correctness), contextually relevant (0.96 on task relevancy), and faithful safety assessments (0.95 and 0.99 on reasoning faithfulness). The results suggest promising performance in construction scene perception, as well as textual analysis and reasoning. This approach represents an advancement in automatic construction safety inspection and contributes to the broader discourse on formalizing multi-modal processing of construction data, offering insights into creating more flexible and comprehensive safety management systems.</div></div>","PeriodicalId":50941,"journal":{"name":"Advanced Engineering Informatics","volume":"65 ","pages":"Article 103246"},"PeriodicalIF":8.0,"publicationDate":"2025-03-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143619131","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}

引用次数: 0

Industrial applications of digital twins: A systematic investigation based on bibliometric analysis

IF 8 1区工程技术 Q1 COMPUTER SCIENCE, ARTIFICIAL INTELLIGENCE

Advanced Engineering Informatics

Pub Date : 2025-03-13 DOI: 10.1016/j.aei.2025.103264

Jiangzhuo Ren , Rafiq Ahmad , Dejun Li , Yongsheng Ma , Jizhuang Hui

Digital twins have evolved into a mature concept, unleashing significant potential across diverse domains. The applications of digital twins are currently experiencing a period of rapid growth, with a particular emphasis on the industrial sector. While previous works have examined the frameworks and architectures of digital twins for industrial applications or explored applications within specific industrial fields, there is a gap in the review work concerning the specific characteristics of digital twins in industrial applications and the industrial processes to which this technology has been applied. To address this gap, digital twins’ overall and industrial perspectives are compared through bibliometric analysis to identify the specific development relationship, research hotspots, and knowledge structure of digital twins in industry. Building upon the bibliometric analysis results, this paper presents a complete survey on the technologies/tools supporting industrial applications. The results indicate that simulation, sensor, and cloud computing are predominant in the basic, core, and advanced technologies. Further, this work investigates various industrial processes utilizing digital twins. By combining the bibliometric analysis, it gives that additive manufacturing and machining processes get more attention from digital twins. Finally, according to Shneider’s theory, the evolution stage of digital twins in the industrial context is analyzed. It may have advanced to the late phase of Stage III, a prolific stage.

{"title":"Industrial applications of digital twins: A systematic investigation based on bibliometric analysis","authors":"Jiangzhuo Ren , Rafiq Ahmad , Dejun Li , Yongsheng Ma , Jizhuang Hui","doi":"10.1016/j.aei.2025.103264","DOIUrl":"10.1016/j.aei.2025.103264","url":null,"abstract":"<div><div>Digital twins have evolved into a mature concept, unleashing significant potential across diverse domains. The applications of digital twins are currently experiencing a period of rapid growth, with a particular emphasis on the industrial sector. While previous works have examined the frameworks and architectures of digital twins for industrial applications or explored applications within specific industrial fields, there is a gap in the review work concerning the specific characteristics of digital twins in industrial applications and the industrial processes to which this technology has been applied. To address this gap, digital twins’ overall and industrial perspectives are compared through bibliometric analysis to identify the specific development relationship, research hotspots, and knowledge structure of digital twins in industry. Building upon the bibliometric analysis results, this paper presents a complete survey on the technologies/tools supporting industrial applications. The results indicate that simulation, sensor, and cloud computing are predominant in the basic, core, and advanced technologies. Further, this work investigates various industrial processes utilizing digital twins. By combining the bibliometric analysis, it gives that additive manufacturing and machining processes get more attention from digital twins. Finally, according to Shneider’s theory, the evolution stage of digital twins in the industrial context is analyzed. It may have advanced to the late phase of Stage III, a prolific stage.</div></div>","PeriodicalId":50941,"journal":{"name":"Advanced Engineering Informatics","volume":"65 ","pages":"Article 103264"},"PeriodicalIF":8.0,"publicationDate":"2025-03-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143619135","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}

引用次数: 0

A Hilbert-based physics-informed neural network for instantaneous meshing frequency estimation of planetary gear set

IF 8 1区工程技术 Q1 COMPUTER SCIENCE, ARTIFICIAL INTELLIGENCE

Advanced Engineering Informatics

Pub Date : 2025-03-12 DOI: 10.1016/j.aei.2025.103250

Shunan Luo , Yinbo Wang , He Dai , Xinhua Long , Zhike Peng

The instantaneous meshing frequency (IMF) is core information for monitoring the operation of planetary gear sets. Due to the complex amplitude modulation, estimating the IMF from vibration signal is challenging. Ridge extraction methods based on time–frequency analysis (TFA) are widely used to estimate the IMF in rotating machinery. However, since these methods process vibration signal in batches, they are unsuitable for online status monitoring applications. In this work, based on the vibration signal model of planetary gear set, a Hilbert-based physics-informed neural network (PINN) is designed to estimate IMF online. The proposed PINN mainly contains three modules. A FIR Hilbert filter is used to extract variation features from vibration signal. An encoder module implemented by transformer network is employed to estimate the IMF. A notch filter group decoder based on vibration signal model is designed to calculate the estimated errors. The parameters of transformer encoder module are updated using error signals. Leveraging the filtering capability of the notch filter group decoder, the PINN adaptively tracks IMF variations without requiring labeled datasets and offline model training. Furthermore, the modules in the PINN process data sequentially, making it well-suited for real-time online status monitoring of planetary gear sets. Simulations and experiments demonstrate the effectiveness and robustness of the PINN for IMF estimation in planetary gear sets.

{"title":"A Hilbert-based physics-informed neural network for instantaneous meshing frequency estimation of planetary gear set","authors":"Shunan Luo , Yinbo Wang , He Dai , Xinhua Long , Zhike Peng","doi":"10.1016/j.aei.2025.103250","DOIUrl":"10.1016/j.aei.2025.103250","url":null,"abstract":"<div><div>The instantaneous meshing frequency (IMF) is core information for monitoring the operation of planetary gear sets. Due to the complex amplitude modulation, estimating the IMF from vibration signal is challenging. Ridge extraction methods based on time–frequency analysis (TFA) are widely used to estimate the IMF in rotating machinery. However, since these methods process vibration signal in batches, they are unsuitable for online status monitoring applications. In this work, based on the vibration signal model of planetary gear set, a Hilbert-based physics-informed neural network (PINN) is designed to estimate IMF online. The proposed PINN mainly contains three modules. A FIR Hilbert filter is used to extract variation features from vibration signal. An encoder module implemented by transformer network is employed to estimate the IMF. A notch filter group decoder based on vibration signal model is designed to calculate the estimated errors. The parameters of transformer encoder module are updated using error signals. Leveraging the filtering capability of the notch filter group decoder, the PINN adaptively tracks IMF variations without requiring labeled datasets and offline model training. Furthermore, the modules in the PINN process data sequentially, making it well-suited for real-time online status monitoring of planetary gear sets. Simulations and experiments demonstrate the effectiveness and robustness of the PINN for IMF estimation in planetary gear sets.</div></div>","PeriodicalId":50941,"journal":{"name":"Advanced Engineering Informatics","volume":"65 ","pages":"Article 103250"},"PeriodicalIF":8.0,"publicationDate":"2025-03-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143601117","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}

引用次数: 0

CFD-guided memory-enhanced LSTM predicts leeward flow of railway windproof structures

IF 8 1区工程技术 Q1 COMPUTER SCIENCE, ARTIFICIAL INTELLIGENCE

Advanced Engineering Informatics

Pub Date : 2025-03-11 DOI: 10.1016/j.aei.2025.103253

Yan-Ke Tan , De-Hui Ouyang , E Deng , Huan Yue , Yi-Qing Ni

Under the protection of windproof structures, wind speed along high-speed railway (HSR) lines decreases while turbulence intensity increases, particularly in transition segments between different windproof structure types. Considering the challenges of placing trackside sensors and the lengthy computational time required for numerical simulations, we propose a novel predictor that integrates computational fluid dynamics (CFD) with neural network (NN) methodologies to estimate internal wind speed signals in real-time based on external measurements from the windproof structures. This architecture combines an echo state network with a dynamic reservoir (dESN) as a feature extractor and a memory-enhanced long short-term memory (eLSTM) network as the output module. It is characterized by a dataset from an experimentally validated delayed detached eddy simulation (DDES)-based CFD model. Applied to a segment of the Lanzhou-Xinjiang railway crossing the Baili windy zone, the predictor achieves over 85% accuracy in estimating in-plane wind speeds and vertical wind profiles, outperforming conventional methods. Additionally, it demonstrates effectiveness under extreme conditions, including exceedingly high incoming flow speeds, absence of downstream measurement points, and sparse or distant sensor arrangements.

{"title":"CFD-guided memory-enhanced LSTM predicts leeward flow of railway windproof structures","authors":"Yan-Ke Tan , De-Hui Ouyang , E Deng , Huan Yue , Yi-Qing Ni","doi":"10.1016/j.aei.2025.103253","DOIUrl":"10.1016/j.aei.2025.103253","url":null,"abstract":"<div><div>Under the protection of windproof structures, wind speed along high-speed railway (HSR) lines decreases while turbulence intensity increases, particularly in transition segments between different windproof structure types. Considering the challenges of placing trackside sensors and the lengthy computational time required for numerical simulations, we propose a novel predictor that integrates computational fluid dynamics (CFD) with neural network (NN) methodologies to estimate internal wind speed signals in real-time based on external measurements from the windproof structures. This architecture combines an echo state network with a dynamic reservoir (dESN) as a feature extractor and a memory-enhanced long short-term memory (eLSTM) network as the output module. It is characterized by a dataset from an experimentally validated delayed detached eddy simulation (DDES)-based CFD model. Applied to a segment of the Lanzhou-Xinjiang railway crossing the Baili windy zone, the predictor achieves over 85% accuracy in estimating in-plane wind speeds and vertical wind profiles, outperforming conventional methods. Additionally, it demonstrates effectiveness under extreme conditions, including exceedingly high incoming flow speeds, absence of downstream measurement points, and sparse or distant sensor arrangements.</div></div>","PeriodicalId":50941,"journal":{"name":"Advanced Engineering Informatics","volume":"65 ","pages":"Article 103253"},"PeriodicalIF":8.0,"publicationDate":"2025-03-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143593409","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}

引用次数: 0

Deformation prediction model for concrete dams considering the effect of solar radiation

IF 8 1区工程技术 Q1 COMPUTER SCIENCE, ARTIFICIAL INTELLIGENCE

Advanced Engineering Informatics

Pub Date : 2025-03-11 DOI: 10.1016/j.aei.2025.103252

Mingkai Liu , Yining Qi , Huaizhi Su

Due to the rarefied atmosphere and shallow cloud layers, high-altitude regions receive greater solar radiation than lower-altitude regions. This intense solar radiation affects the durability and temperature field of concrete dams, thereby influencing the deformation behavior. A deformation prediction model for concrete dams is developed that considers the impact of solar radiation in this study. Initially, the Bayesian online changepoint detection algorithm, coupled with the density-based spatial clustering of applications with noise algorithm, is employed to analyze the solar radiation data for two concrete dams located at the same latitude but differing in altitude, to assess potential disparities. Subsequently, based on the principles of heat transfer and the absorption efficiency of solar radiation, the impact of solar radiation is quantified, thereby refining the input factors of the proposed model. Finally, by incorporating the Multi-Head Self-Attention mechanism into the Long Short-Term Memory model, deformation data prediction is achieved, and the attention weights are output to deeply analyze the impact of different input factors on the deformation magnitude. An engineering case study serves to validate the practical applicability of the proposed model. The case analysis results highlight significant differences in solar radiation data between high-altitude and low-altitude regions and show that accounting for the impact of solar radiation can effectively enhance the performance of the prediction model.

{"title":"Deformation prediction model for concrete dams considering the effect of solar radiation","authors":"Mingkai Liu , Yining Qi , Huaizhi Su","doi":"10.1016/j.aei.2025.103252","DOIUrl":"10.1016/j.aei.2025.103252","url":null,"abstract":"<div><div>Due to the rarefied atmosphere and shallow cloud layers, high-altitude regions receive greater solar radiation than lower-altitude regions. This intense solar radiation affects the durability and temperature field of concrete dams, thereby influencing the deformation behavior. A deformation prediction model for concrete dams is developed that considers the impact of solar radiation in this study. Initially, the Bayesian online changepoint detection algorithm, coupled with the density-based spatial clustering of applications with noise algorithm, is employed to analyze the solar radiation data for two concrete dams located at the same latitude but differing in altitude, to assess potential disparities. Subsequently, based on the principles of heat transfer and the absorption efficiency of solar radiation, the impact of solar radiation is quantified, thereby refining the input factors of the proposed model. Finally, by incorporating the Multi-Head Self-Attention mechanism into the Long Short-Term Memory model, deformation data prediction is achieved, and the attention weights are output to deeply analyze the impact of different input factors on the deformation magnitude. An engineering case study serves to validate the practical applicability of the proposed model. The case analysis results highlight significant differences in solar radiation data between high-altitude and low-altitude regions and show that accounting for the impact of solar radiation can effectively enhance the performance of the prediction model.</div></div>","PeriodicalId":50941,"journal":{"name":"Advanced Engineering Informatics","volume":"65 ","pages":"Article 103252"},"PeriodicalIF":8.0,"publicationDate":"2025-03-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143593234","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}

引用次数: 0