Computers in Industry最新文献

{"title":"Enhancing retrieval-augmented generation for interoperable industrial knowledge representation and inference toward cognitive digital twins","authors":"Dachuan Shi ,&nbsp;Jianzhang Li ,&nbsp;Olga Meyer ,&nbsp;Thomas Bauernhansl","doi":"10.1016/j.compind.2025.104330","DOIUrl":"10.1016/j.compind.2025.104330","url":null,"abstract":"<div><div>The escalating volume and complexity of digital data within the manufacturing sector highlight an urgent need for an efficient knowledge representation and inference solution. Traditional approaches, which often rely on ontologies, knowledge graphs, or digital twins (DTs) for knowledge representation, and rule-based algorithms for inference, are becoming insufficient. The emergence of generative AI, particularly large language models (LLM) and retrieval-augmented generation (RAG), offers a more efficient and intelligent alternative. However, the performance of an RAG system is heavily dependent on the quality of retrieval results, which can be compromised by domain-specific knowledge and retrieval distractors. To address this challenge, we propose to enhance RAG systems tailored for the manufacturing industry in two aspects. First, we utilize the Asset Administration Shell (AAS), which represents the German industrial perspective on cognitive DTs, to create a representation of assets and knowledge in standardized information models. This establishes a robust foundation for the retrieval sources. Second, we propose a contrastive selection loss (CSL) to fine-tune an open-source LLM to refine the retrieval results. Fine-tuned LLMs possess higher efficiency and accuracy on task- and domain-specific datasets, while the CSL further enhances the model's ability to distinguish true positives from similar distractors. The enhanced RAG system is demonstrated in a robotic work cell integration use case and evaluated through a novel evaluation protocol. Additionally, the retrieval effectiveness of the RAG system, specifically the LLM fine-tuned with CSL, is extensively validated through statistical experiments. The results confirm its superior performance over state-of-the-art methods, including GPT-4 with in-context learning prompts and other fine-tuned models.</div></div>","PeriodicalId":55219,"journal":{"name":"Computers in Industry","volume":"171 ","pages":"Article 104330"},"PeriodicalIF":8.2,"publicationDate":"2025-06-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144480629","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Vision-based hand pose estimation methods for Augmented Reality in industry: Crowdsourced evaluation on HoloLens 2","authors":"Kamil Żywanowski ,&nbsp;Mikołaj Łysakowski ,&nbsp;Michał R. Nowicki ,&nbsp;Jason T. Jacques ,&nbsp;Sławomir K. Tadeja ,&nbsp;Thomas Bohné ,&nbsp;Piotr Skrzypczyński","doi":"10.1016/j.compind.2025.104328","DOIUrl":"10.1016/j.compind.2025.104328","url":null,"abstract":"<div><div>Gestural input based on hand pose estimation is a common interaction method for augmented reality (AR). This interaction technique has gained more popularity with the emergence of novel AR-supporting devices such as Microsoft HoloLens 2 (HL2) and advancements in computer vision research underpinning hand-tracking and gesture recognition methods. In our work, we focus on challenging cases where the AR interface is facilitated with a state-of-the-art HL2 headset for unconstrained execution of tasks requiring simultaneous hand movement and tracking. When using this headset, AR users might bimanually interact with digital and physical objects that are visible in the user’s field of view (FoV) through the see-through visor. Due to the limiting in-built capabilities, we investigated a range of hand pose estimation functionalities from different domains. To ensure a fair comparison, we asked several participants to carry out tasks requiring interactions with real-world objects and record the performance of various hand-tracking solutions. Next, we evaluated the performance of these algorithms through crowdsourcing, often used to provide ground truth for machine learning training. Our results provide a guideline for AR developers in selecting appropriate hand-tracking solutions for a given deployment context.</div></div>","PeriodicalId":55219,"journal":{"name":"Computers in Industry","volume":"171 ","pages":"Article 104328"},"PeriodicalIF":8.2,"publicationDate":"2025-06-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144365666","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Energy consumption forecasting in non-stationary machining processes based on a physics-guided transformer","authors":"Meihang Zhang ,&nbsp;Ruiping Wang","doi":"10.1016/j.compind.2025.104321","DOIUrl":"10.1016/j.compind.2025.104321","url":null,"abstract":"<div><div>Predicting energy consumption in non-stationary machining processes is challenging due to data complexity, dynamic variations, and real-time requirements. This paper proposes a novel physics-guided Transformer model incorporating supervisory-compensatory mechanisms. Firstly, several data preprocessing and feature extraction techniques, including Lagrange interpolation, wavelet transform, and a combined approach of principal component analysis and correlation analysis, are employed to enhance data quality and identify key physical variable. Secondly, modeling accuracy and efficiency are improved by integrating physics-guided variables into the traditional Transformer model, leading to advancements in the existing de-stationary attention module. Finally, distinct training and prediction models are established, incorporating a self-supervised, self-compensation mechanism in the training phase. This mechanism utilizes ground truth for training convergence, with optimized model parameters subsequently applied to the prediction model, significantly enhancing predictive efficacy. Experimental results demonstrate that the proposed method outperforms state-of-the-art approaches, achieving improvements in energy consumption prediction accuracy of over 76 % for carbon fiber processing, 30 % for plastic, 32.7 % for aluminum, and 54.5 % for 45steel. The integration of physical principles with sequence modeling enables precise energy consumption forecasting in non-stationary machining, advancing both predictive accuracy and industrial energy management.</div></div>","PeriodicalId":55219,"journal":{"name":"Computers in Industry","volume":"171 ","pages":"Article 104321"},"PeriodicalIF":8.2,"publicationDate":"2025-06-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144306358","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"MiniMaxAD: A lightweight autoencoder for feature-rich anomaly detection","authors":"Fengjie Wang,&nbsp;Chengming Liu,&nbsp;Lei Shi,&nbsp;Haibo Pang","doi":"10.1016/j.compind.2025.104315","DOIUrl":"10.1016/j.compind.2025.104315","url":null,"abstract":"<div><div>Previous industrial anomaly detection (IAD) methods often struggle to handle the extensive diversity in training sets, particularly when they contain stylistically diverse and feature-rich samples, which we categorize as feature-rich anomaly detection datasets (FRADs). This challenge is evident in applications such as multi-view and multi-class scenarios. To address this challenge, we developed MiniMaxAD, a efficient autoencoder designed to efficiently compress and memorize extensive information from normal images. Our model employs a technique that enhances feature diversity, thereby increasing the effective capacity of the network. It also utilizes large kernel convolution to extract highly abstract patterns, which contribute to efficient and compact feature embedding. Moreover, we introduce an Adaptive Contraction Hard Mining Loss (ADCLoss), specifically tailored to FRADs. In our methodology, any dataset can be unified under the framework of feature-rich anomaly detection, in a way that the benefits far outweigh the drawbacks. Our approach has achieved state-of-the-art performance in multiple challenging benchmarks. Code is available at: <span><span>https://github.com/WangFengJiee/MiniMaxAD</span><svg><path></path></svg></span>.</div></div>","PeriodicalId":55219,"journal":{"name":"Computers in Industry","volume":"171 ","pages":"Article 104315"},"PeriodicalIF":8.2,"publicationDate":"2025-06-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144212331","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Autonomous vehicle crash risk modeling by integrating data augmentation and two-layer stacking","authors":"Leipeng Zhu ,&nbsp;Zhiqing Zhang ,&nbsp;Yongnan Zhang ,&nbsp;Jingyang Yu ,&nbsp;Hongjia Wang","doi":"10.1016/j.compind.2025.104320","DOIUrl":"10.1016/j.compind.2025.104320","url":null,"abstract":"<div><div>Autonomous vehicle (AV) technology aims to eliminate traffic crashes caused by driver errors, but its adoption has introduced new types of crashes. Due to the high dimensionality and limited sample size of AV crash data, identifying underlying risk factors remains challenging, and crash predictive performance is often suboptimal. To address these issues, this study develops an interpretable data augmentation strategy and the optimized two-layer stacking algorithm, further integrating them into a unified framework that accurately identifies key crash contributing factors and significantly improves predictive performance. The findings reveal that: 1) AV crashes show significant variation in their temporal distributions but follow consistent spatial agglomeration patterns. 2) AV reliability significantly decreases in high-interaction scenarios, with peak travel times and uncertain road conditions identified as key contributing factors. 3) The data augmentation algorithm enhances on key contributing factors and the feature crosses, enhances the model’s ability to capture nonlinear relationships in crash data and improves predictive accuracy in small-sample scenarios, particularly for injury-related crashes. 4) The optimized two-layer stacking algorithm integrates the heterogeneous learning capabilities of models such as LightGBM and Random Forest, significantly improving the ability to recognize complex crash patterns. When combined with data augmentation, the framework achieves strong predictive performance, with both precision and recall reaching 0.92 and the area under the receiver operating characteristic curve at 0.96. Compared to existing machine learning approaches, this framework shows notable advantages in handling high-dimensional small-sample AV crash data. The framework provides an effective solution for AV crash risk modeling and safety design, contributing to the development and implementation of safer intelligent transportation systems.</div></div>","PeriodicalId":55219,"journal":{"name":"Computers in Industry","volume":"171 ","pages":"Article 104320"},"PeriodicalIF":8.2,"publicationDate":"2025-06-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144195780","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Industrial Internet of Things: Implementations, challenges, and potential solutions across various industries","authors":"Shaila Afrin ,&nbsp;Sabiha Jannat Rafa ,&nbsp;Maliha Kabir ,&nbsp;Tasfia Farah ,&nbsp;Md. Sakib Bin Alam ,&nbsp;Aiman Lameesa ,&nbsp;Shams Forruque Ahmed ,&nbsp;Amir H. Gandomi","doi":"10.1016/j.compind.2025.104317","DOIUrl":"10.1016/j.compind.2025.104317","url":null,"abstract":"<div><div>The Industrial Internet of Things (IIoT) has emerged as a potent catalyst for transformation across many industries as a part of Industry 4.0. This review thoroughly examines IIoT applications, demonstrating how it enhances operational efficiency, informed decision-making, cost optimization, innovation, and workplace safety. While prior research has often concentrated on technical dimensions such as fog and edge computing, network protocols, or big data integration, several emerging and high-impact application areas remain underexplored. This study addresses that gap by systematically reviewing IIoT implementations in critical yet often overlooked domains, including environmental monitoring, agriculture, construction, healthcare, robotics, smart grids, and predictive maintenance. It offers fresh insights into how IIoT is being adapted to meet real-world challenges in these sectors. In addition to outlining the current landscape, the review identifies core barriers such as data security, interoperability, and system scalability. It underscores the importance of cross-sector collaboration and strategic alignment to fully leverage the transformative potential of IIoT. The paper concludes by outlining key research gaps and future opportunities to guide continued innovation and scholarly investigation.</div></div>","PeriodicalId":55219,"journal":{"name":"Computers in Industry","volume":"170 ","pages":"Article 104317"},"PeriodicalIF":8.2,"publicationDate":"2025-05-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144169086","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"SRLFormer: Single Retinex-based and low-light image guidance Transformer for low-light image enhancement","authors":"Bin Wang,&nbsp;Bini Zhang,&nbsp;Jinfang Sheng","doi":"10.1016/j.compind.2025.104314","DOIUrl":"10.1016/j.compind.2025.104314","url":null,"abstract":"<div><div>In image enhancement for low-illumination images, deep learning methods based on the Retinex theory typically decompose the image into illumination and reflectance, followed by iterative optimization or the use of prior custom enhancements. The reflectance map is then approximated as the enhanced image by dividing the radiance by the illumination map. However, this approach does not account for the noise hidden in low-illumination images or introduced during the enhancement of illumination. Additionally, it may cause computational overflow and amplify noise when the illumination in certain regions approaches ”0”. Moreover, these methods often require cumbersome multi-stage training and rely solely on convolutional neural networks, indicating limitations in capturing long-range dependencies. This paper proposes an efficient single-stage framework named SRF(Retinex-based single-retinex-based framework based on Retinex). SRF first estimates the inverse illumination image, then enhances the image by multiplying the inverse illumination with the low-illumination image, resulting in an image with improved brightness but still containing noise. Finally, we design a low-illumination guided Transformer network, LGF (Low-Illumination Guided Transformer), which utilizes the low-illumination image to guide denoising, thus more comprehensively considering the edge and detail information of the enhanced image. By integrating the LGT into SRF, we obtain the proposed algorithm SRLFormer. Experimental results show that SRLFormer significantly outperforms state-of-the-art methods in both qualitative and quantitative experiments, and its potential practical value is also demonstrated in downstream tasks and applications.</div></div>","PeriodicalId":55219,"journal":{"name":"Computers in Industry","volume":"170 ","pages":"Article 104314"},"PeriodicalIF":8.2,"publicationDate":"2025-05-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144139286","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"A multiscale process-aware retention network for fault prediction in mixed-model production","authors":"Mingda Chen ,&nbsp;Ruiyun Yu ,&nbsp;Zhiyuan Liang ,&nbsp;Kun Li ,&nbsp;Haifei Qi","doi":"10.1016/j.compind.2025.104313","DOIUrl":"10.1016/j.compind.2025.104313","url":null,"abstract":"<div><div>In the manufacturing industry, the demand for fault-prediction solutions is increasing to prevent unexpected downtimes and reduce maintenance costs. Although deep-learning methods have demonstrated excellent performance in this domain, the current methods typically overlook the analysis of variable and random processes within mixed-model production, which is a manufacturing strategy that offers flexibility and efficiency in satisfying diverse consumer demands. Hence, we propose the multiscale process-aware retention network (MPRNet), which segments a time series into multiscale patches, thus enabling the model to focus on local information within each production process and correlations across all production processes. Furthermore, the network incorporates a cross-channel interaction module designed to dynamically capture the interactions between various sensor data types using a graph attention network, as well as transmit fault information across processes using state equations. We validate our proposed model on the BBA stud welding gun dataset and four additional open case studies. Compared with other established fault-prediction and time-series models, the MPRNet demonstrates improved F1-score by 13.1% in the BBA case and consistently achieves the best or near-best results in the open case studies.</div></div>","PeriodicalId":55219,"journal":{"name":"Computers in Industry","volume":"170 ","pages":"Article 104313"},"PeriodicalIF":8.2,"publicationDate":"2025-05-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144139287","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Multi-style adversarial variational self-distillation in randomized domains for single-domain generalized fault diagnosis","authors":"Fan Yang,&nbsp;Xiaofeng Liu,&nbsp;Chunbing Zhang,&nbsp;Lin Bo","doi":"10.1016/j.compind.2025.104319","DOIUrl":"10.1016/j.compind.2025.104319","url":null,"abstract":"<div><div>As rotating machinery often operates under complex and variable harsh conditions, domain generalization-based fault diagnosis has been adopted to tackle the challenge of distribution shifts and unseen data in target domains. However, most existing methods depend on fully labeled data from multiple source domains to learn domain-invariant representations. In practice, collecting comprehensive labeled data across diverse working conditions is often impractical, resulting in data insufficiency and distribution inconsistencies. To address the challenging scenario in which only a single fully labeled source domain is available, this article proposes a multi-style adversarial variational self-distillation (MSAVSD) framework based on domain randomization for single-domain generalized fault diagnosis. First, a domain-randomized generation module is developed to adaptively generate samples following randomized distributions by integrating adaptive noise and multi-scale style learning, thereby enriching the synthetic data with diverse and informative fault representations. Next, a scale-enhanced feature extraction module is introduced to extract rich domain-invariant features, thereby maximizing the utilization of fault-related information under limited training conditions. The method suppresses task-irrelevant noise and redundancy via variational self-distillation and employs contrastive learning to enhance the discriminability and consistency of task-relevant features. Extensive diagnostic experiments on three datasets, two self-collected and one publicly available, demonstrate that the proposed method outperforms state-of-the-art methods.</div></div>","PeriodicalId":55219,"journal":{"name":"Computers in Industry","volume":"170 ","pages":"Article 104319"},"PeriodicalIF":8.2,"publicationDate":"2025-05-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144131011","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Survey of automated methods for design and assessment of smart products","authors":"Anoop Kumar Sinha ,&nbsp;Youngmi Christina Choi ,&nbsp;David W. Rosen","doi":"10.1016/j.compind.2025.104316","DOIUrl":"10.1016/j.compind.2025.104316","url":null,"abstract":"<div><div>User centric smart products prioritize the needs and preferences of users, enhancing their experience and satisfaction. Involving users in the design and assessment of smart products ensures that they meet real-world requirements, leading to more intuitive product design, user interface, and functionalities that truly resonate with users. Further, the capability of generating and evaluating many alternative designs early in product development is beneficial. However, the need to construct physical prototypes for user testing limits the number of designs that can be evaluated during early design stages. As such, our interest is in <u>automated</u> methods that support user centered design and usability and user experience assessment. In this review article, we look at at two decades of automation methods that have been employed in the design and development of user centric smart products. The focus of these automation methods is to incorporate user voice in early design stages rather than replacing the users. We have identified five key activities of the design cycle in which automated methods have been employed: design thinking, design ideation, prototype creation, user data collection for usability study, and user data analysis. Overall, 154 articles were identified across engineering, human-computer interaction, human factors, inclusive design, industrial design, and other disciplines that have incorporated automation methods to include the user’s voice in the design of user centric smart products. This review examines the effectiveness and limitations of different automation methods compared to conventional methods, offering valuable insights and suggestions to enhance the design processes of smart products with a focus on widespread usability issues. Our specific interest lies in developing assistive mobility and rehabilitation devices, where constraints such as limited development time and resources persist, yet the usability and user experience profoundly influence significant outcomes like perceived functionality, stigma, and device acceptance.</div></div>","PeriodicalId":55219,"journal":{"name":"Computers in Industry","volume":"170 ","pages":"Article 104316"},"PeriodicalIF":8.2,"publicationDate":"2025-05-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144131013","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}