Journal of Manufacturing Systems最新文献

{"title":"A frequency mask and decoupling max-logit based XAI method to explain DNN for fault diagnosis","authors":"Junfei Du,&nbsp;Yiping Gao,&nbsp;Liang Gao,&nbsp;Xiuyu Li","doi":"10.1016/j.jmsy.2025.06.004","DOIUrl":"10.1016/j.jmsy.2025.06.004","url":null,"abstract":"<div><div>Recently, various deep neural network (DNN) models have been proposed for fault diagnosis. Owing to the black-box nature of the DNN, Diagnosis results are unexplainable. Therefore, explainable artificial intelligence (XAI) methods are required. However, it is difficult for existing XAI methods to separate fault and irrelevant features because the fault features are instantaneous. To address this issue, a frequency mask and decoupling max-logit-based XAI method (FM-Explainer) is proposed to explain the DNN for fault diagnosis. Because the fault features can be well represented in the frequency domain, the proposed method optimizes a mask on the frequency domain of the input to identify the fault features. In addition, to avoid unreliable explanations caused by out-of-distribution (OoD) data, a regularization is designed based on decoupling max-logit, and the spatial penalty is used, which ensures that no irrelevant features remain in the explanation. Extensive experiments are carried out to verify the effectiveness of the proposed method using five quantitative evaluation metrics: Insertion/Deletion, Sensitivity-N, and Degradation. The results show that the FM-Explainer outperforms existing methods, and explanations by the FM-Explainer are consistent with the fault characteristic frequency. This indicates that the FM-Explainer is effective in precisely identifying fault features.</div></div>","PeriodicalId":16227,"journal":{"name":"Journal of Manufacturing Systems","volume":"82 ","pages":"Pages 98-113"},"PeriodicalIF":12.2,"publicationDate":"2025-06-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144254010","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":"Energy flexible manufacturing systems: A systematic literature review, trends and challenges","authors":"Longyao Xu ,&nbsp;Peiji Liu ,&nbsp;Xu Wang ,&nbsp;Yifei Lin ,&nbsp;Yameng Shi ,&nbsp;Xi Huang ,&nbsp;Xi Vincent Wang","doi":"10.1016/j.jmsy.2025.06.010","DOIUrl":"10.1016/j.jmsy.2025.06.010","url":null,"abstract":"<div><div>Utilizing renewable energy nearby has become a critical measure for manufacturing to reduce carbon emissions and energy cost. To mitigate the impact of fluctuations in renewable energy on manufacturing, energy flexibility (EF) technology has become a key approach to balance energy supply and demand. Furthermore, with deepening application of EF technology in manufacturing systems, energy flexible manufacturing systems (EFMS) have gradually become a promising direction for the future development of manufacturing systems. However, since EFMS involves multiple fields, the current understanding of EFMS remains unclear. In this review, we provide a comprehensive overview of the advancements, trends and challenges in EF technology and EFMS. For this, this paper first summarizes the definitions and evaluation indicators of EF in different fields; Second, a quantitative analysis was carried out on 68 studies related to EF technology and EFMS from 5 aspects; Beyond the overview of literature, we identify the trends and challenges for EFMS. On the one hand, the trends of EFMS include low carbonization, digitalization, intelligentization, flexibility and clustering. On the other hand, identified challenges encompass aspects like energy-related data processing, energy flow modeling, prediction, regulation and system design. Through this work, it is hoped to provide comprehensive guidance for addressing common challenges in the development of EFMS from diverse research perspectives, while highlighting potential future research trends and challenges.</div></div>","PeriodicalId":16227,"journal":{"name":"Journal of Manufacturing Systems","volume":"82 ","pages":"Pages 83-97"},"PeriodicalIF":12.2,"publicationDate":"2025-06-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144254011","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":"Towards industry 5.0: A framework of reconfigurable matrix-structured manufacturing system","authors":"Miao Wang ,&nbsp;Yifei Tong ,&nbsp;Cunbo Zhuang ,&nbsp;Xiaodong Du","doi":"10.1016/j.jmsy.2025.06.002","DOIUrl":"10.1016/j.jmsy.2025.06.002","url":null,"abstract":"<div><div>Current manufacturing faces increasingly volatile supply chains and diverse customer demands, requiring a balance between flexibility and efficiency. The matrix-structured manufacturing system (MMS) is a promising solution to frequent disruptions, yet existing research lacks a comprehensive framework for real-world implementation. This study proposes a reconfigurable matrix-structured manufacturing system (RMMS) that integrates advanced Industry 4.0 (I4.0) digital tools with Industry 5.0 (I5.0) visions of resilience and human–centricity. By introducing a “configuration” concept, RMMS organizes logical production lines to address the randomness and uncertainty of workflow typically seen in classic MMS, enabling efficient multi-variety, variable-batch production. Building on this foundation, we present the conceptual and systematic architecture of RMMS. Besides, a feasible technology roadmap is given, including order feature analysis, cell formation method and multidimensional reconfiguration mechanism. To validate our approach, we demonstrate the simulation comparisons and software implementation of RMMS through a case study in a high-precision electronics technology company, which shows improved resilience and rapid changeovers. Overall, the results indicate that RMMS provides a practical blueprint for human-centric, resilient manufacturing in dynamic market conditions, effectively bridging the gaps in current MMS implementations and advancing the evolving I5.0 landscape.</div></div>","PeriodicalId":16227,"journal":{"name":"Journal of Manufacturing Systems","volume":"82 ","pages":"Pages 114-136"},"PeriodicalIF":12.2,"publicationDate":"2025-06-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144263138","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":"Physics-informed orthogonal network with hierarchical time-frequency feature refining strategy for tool wear recognition","authors":"Yujun Zhou,&nbsp;Tangbin Xia,&nbsp;Rourou Li,&nbsp;Yuhui Xu,&nbsp;Guojin Si,&nbsp;Lifeng Xi","doi":"10.1016/j.jmsy.2025.06.009","DOIUrl":"10.1016/j.jmsy.2025.06.009","url":null,"abstract":"<div><div>Tool wear recognition is critical to improve the safety and reliability of machining operations with real-time tool status assessment. Conventional deep learning-based (DL-based) recognition approaches map time-frequency representations (TFRs) of the monitoring signals to tool wear with neural networks. However, data-driven mapping suffers from hardships in excavating wear-sensitive information due to the lack of explicit constraints on wear mechanisms, resulting in inferior recognition performance and recognition results against physical laws. To address this issue, this paper develops a time-frequency refining physics-informed orthogonal network (TFRPION). Firstly, a hierarchical time-frequency refining strategy consisting of energy concentration and adaptive amplitude modulation is conducted to emphasize machining dynamics-related characteristic frequency components in TFRs, highlighting wear-sensitive signal features. Secondly, an orthogonal network module maps features from the refined TFRs by capturing temporal amplitude variations within characteristic frequency components, improving the physical representational capability for the TFR mapping process. Thirdly, a physical network module and a modified loss function that take wear mechanisms into account are integrated to regularize the calculation path and optimization of the proposed network, enhancing the physical consistency between its mapping process and wear mechanisms. The feasibility and effectiveness of the proposed network are verified with collected spindle current signals in high-speed milling tool wear recognition experiments.</div></div>","PeriodicalId":16227,"journal":{"name":"Journal of Manufacturing Systems","volume":"82 ","pages":"Pages 60-82"},"PeriodicalIF":12.2,"publicationDate":"2025-06-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144254012","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":"Comprehensive issue identification for manufacturing data analytics implementation: Systematic literature review and case studies","authors":"Sa-Eun Park ,&nbsp;Sang-Jae Lee ,&nbsp;Hyerim Bae ,&nbsp;Ki-Hun Kim ,&nbsp;Eung-Jun Kang ,&nbsp;Jae-Sung Kim ,&nbsp;Yong-Min Park ,&nbsp;Min-Ji Park","doi":"10.1016/j.jmsy.2025.05.006","DOIUrl":"10.1016/j.jmsy.2025.05.006","url":null,"abstract":"<div><div>Manufacturers are increasingly adopting manufacturing data analytics (MDA) as a key factor for smart manufacturing. However, successful MDA implementation remains limited due to various issues. Existing studies have barely suggested a comprehensive such issues by focusing on parts of technological, organizational, and environmental (TOE) contexts or issues limited to partial steps of MDA. This study addresses these gaps by identifying comprehensive issue set for MDA implementation (CISM) through a systematic review of 35 papers. The 29 distinct issues with 9 categories were derived to cover both TOE contexts and the five major steps of MDA. The comprehensiveness of CISM was validated through three real-world MDA implementation case studies. CISM is expected to suggest issues for manufacturers to address proactively in MDA implementation and to serve as a basis for stimulating future studies on MDA implementation.</div></div>","PeriodicalId":16227,"journal":{"name":"Journal of Manufacturing Systems","volume":"82 ","pages":"Pages 42-59"},"PeriodicalIF":12.2,"publicationDate":"2025-06-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144243209","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":"Enhancing yield and process efficiency through dual internet of things and augmented reality for AI-driven human-machine interaction in centering mass production","authors":"Yu-Chen Liang,&nbsp;Shiau-Cheng Shiu,&nbsp;Chun-Wei Liu","doi":"10.1016/j.jmsy.2025.06.003","DOIUrl":"10.1016/j.jmsy.2025.06.003","url":null,"abstract":"<div><div>With the emergence of smart manufacturing, artificial intelligence (AI) has become a pivotal technology for enhancing industrial process efficiency and production yield. By integrating data analysis methods, AI can effectively capture process characteristics during manufacturing. However, tasks such as setting machine parameters still rely heavily on human expertise. This study focused on the centering process of optical glass lenses as a case study. To minimize dependence on human expertise, establish real-time diagnostic mechanisms, and shorten calibration times, an intelligent human-machine interactive manufacturing system featuring a Dual Internet of Things (Dual-IoT) architecture and augmented reality (AR) technology was developed. This system employs a feature extraction model that combines root mean square (RMS) with exponentially weighted moving average (EWMA) to analyze time-series signals during processing. Subsequently, an echo state network (ESN) prediction model was established to accurately forecast real-time signals and identify anomalies. In this setup, the control system and AI model are interconnected through a Dual-IoT architecture, enabling real-time data transmission to the intelligent AR-based human-machine interaction system and remote monitoring interface. This setup enables the visualization of process diagnostics and decision-making, providing feedback to the centering machine through remote control mechanisms. According to the verification results, at target specifications of &lt; 0.01 mm roundness and &lt;E0.5 edge cracks, the proposed system enhanced production yield from 64 % to 94 % while reducing production time by 29.2 %. These results confirm the system’s effectiveness in augmenting industrial production processes.</div></div>","PeriodicalId":16227,"journal":{"name":"Journal of Manufacturing Systems","volume":"82 ","pages":"Pages 26-41"},"PeriodicalIF":12.2,"publicationDate":"2025-06-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144243208","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":"Environmental impact of powder production for additive manufacturing: Carbon footprint and cumulative energy demand of gas atomization","authors":"Svenja Ehmsen ,&nbsp;Janosch Conrads ,&nbsp;Matthias Klar ,&nbsp;Jan C. Aurich","doi":"10.1016/j.jmsy.2025.05.004","DOIUrl":"10.1016/j.jmsy.2025.05.004","url":null,"abstract":"<div><div>The production of metal powder required for certain metal additive manufacturing processes has a significant environmental impact on the process chain. In particular, there is a lack of energy- and resource-related data on the environmental impact of industrial powder production and in-depth analysis of individual process steps. This study aims to provide a reliable life cycle inventory and, based on this, to determine the global warming potential (GWP) and cumulative energy demand (CED) resulting from the industrial production of melt atomized metal powders for additive manufacturing using gas atomization within the framework of a life cycle assessment (LCA). In this LCA, considering an average electricity mix at a production site in Germany, the GWP for closed-coupled atomization ranged from 4.61 kg CO₂-eq./kg to 16.71 kg CO₂-eq./kg. The results are slightly lower than those of free-fall atomization with a GWP between 5.58 kg CO₂-eq./kg and 24.81 kg CO₂-eq./kg. The need for inert gas is a major contributor to the environmental impact. If argon is used as an atomizing gas instead of nitrogen, the environmental impact increases, since argon has a GWP and CED approximately six times higher than nitrogen. Preheating the inert gas reduces the requirement and thus also the resulting environmental impact. This study provides a crucial basis for assessing the environmental impact of powder metal additive manufacturing processes and, enabling environmentally friendly process and product design. In addition, effective strategies to reduce the environmental impact of gas atomization can be identified based.</div></div>","PeriodicalId":16227,"journal":{"name":"Journal of Manufacturing Systems","volume":"82 ","pages":"Pages 13-25"},"PeriodicalIF":12.2,"publicationDate":"2025-06-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144221332","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":"A Digital Twin-based condition monitoring system to detect and resolve web slip at traction rollers in a web processing machine","authors":"Arul K. Mathivanan,&nbsp;Jeroen D.M. De Kooning,&nbsp;Kurt Stockman","doi":"10.1016/j.jmsy.2025.05.001","DOIUrl":"10.1016/j.jmsy.2025.05.001","url":null,"abstract":"<div><div>Slippage of web material over rollers is an undesirable phenomenon in web processing applications, causing damage to the web material. This leads to compromised quality and increased waste. While web slippage is commonly observed at high web speeds due to air entrapment on freely rotating rollers, it also occurs at lower web speeds at the traction rollers, which are designed to drive the web through the machine. Installing web speed sensors to detect such web slippage on multiple traction rollers in large-scale web processing machines is expensive. This work presents a novel Digital Twin methodology for online condition monitoring and fault detection to identify web slip at the traction rollers. The Digital Twin uses the contact forces between the web and the traction roller to detect web slippage, greatly reducing the need for web speed sensors and thereby cutting costs and time. Additionally, in response to detected web slip, the newly proposed Digital Twin further acts to resolve the slip. Experimental results demonstrate the effectiveness of the Digital Twin in detecting and resolving web slippage on three different web materials.</div></div>","PeriodicalId":16227,"journal":{"name":"Journal of Manufacturing Systems","volume":"82 ","pages":"Pages 1-12"},"PeriodicalIF":12.2,"publicationDate":"2025-06-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144195179","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":"Interpretable deep temporal neural networks for in-situ monitoring under varying conditions in micro-electrical discharge machining","authors":"Long Ye ,&nbsp;Cheng Guo ,&nbsp;Ming Wu ,&nbsp;Jun Qian ,&nbsp;Nan Yu ,&nbsp;Dominiek Reynaerts","doi":"10.1016/j.jmsy.2025.05.007","DOIUrl":"10.1016/j.jmsy.2025.05.007","url":null,"abstract":"<div><div>Monitoring is critical enabler of digitalization in modern manufacturing, supporting enhanced process control, quality assurance, and real-time decision-making. By integrating data-driven techniques with the powerful capabilities of deep learning, monitoring systems can efficiently extract valuable insights from complex, high-dimensional time-series data. However, traditional data-driven approaches often lack interpretability, limiting their adoption in industrial applications that demand high reliability and accountability. To address this challenge, this paper proposes an interpretable monitoring framework based on a deep temporal neural network (DTNN). Designed with a modular architecture, the DTNN integrates key components for embedding, temporal feature learning and classification, enabling it to effectively capture complex underlying patterns in temporal process data and overcome the limitations of conventional methods. The DTNN’s capabilities are demonstrated in the context of micro-electrical discharge machining (micro-EDM), a prominent non-traditional machining process known for producing intricate and high-precision components. Through a pulse discrimination task utilizing a large dataset of reliable labels, the DTNN achieves superior classification accuracy under varying processing parameters while providing interpretable insights into discharge phenomena. Furthermore, the DTNN monitoring approach is applied to a deep-hole drilling process in micro-EDM, enabling closed-loop control of discharge status and ensuring long-term process stability. The DTNN’s modular design, interpretability and real-time adaptability underscore its potential for advancing data-driven monitoring systems in digital manufacturing.</div></div>","PeriodicalId":16227,"journal":{"name":"Journal of Manufacturing Systems","volume":"81 ","pages":"Pages 222-237"},"PeriodicalIF":12.2,"publicationDate":"2025-05-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144184398","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":"A method for monitoring machining errors of complex thin-walled parts based on the fusion of physical information and CNN","authors":"Wangfei Li ,&nbsp;Junxue Ren ,&nbsp;Kaining Shi ,&nbsp;Yanru Lu ,&nbsp;Huan Zheng","doi":"10.1016/j.jmsy.2025.05.017","DOIUrl":"10.1016/j.jmsy.2025.05.017","url":null,"abstract":"<div><div>Machining errors of complex thin-walled parts have a direct impact on product quality and performance, making their monitoring essential. The monitoring of machining errors of such parts often depends on relevant physical information. However, time-varying physical information (TVPI) is influenced by the dynamic response of the measurement system, and the spatial dynamic relationship between the physical information and machining errors is highly intricate, posing significant challenges for monitoring. To address these challenges, a method based on the fusion of physical information and Convolutional Neural Network (CNN) is proposed for monitoring machining errors of complex thin-walled parts. Initially, a TVPI identification method based on physical theory is introduced, and the spectrum amplitudes of cutting forces are extracted as the TVPI for monitoring machining errors. The feature extraction and nonlinear regression modeling capabilities of the CNN are then leveraged to filter the physical information intelligently and learn the complex relationship between the physical information and machining errors. Ultimately, a monitoring method for machining errors based on the fusion of physical information and the CNN is proposed and experimentally validated on complex thin-walled parts such as blades. Compared with traditional feature identification methods, the TVPI identification method provides enhanced physical interpretability. Additionally, the fusion of physical information and the CNN notably improves the monitoring performance. Compared with monitoring methods based on Gaussian Process Regression, Deep Neural Network and Long Short-Term Memory, the monitoring method based on the fusion of physical information and the CNN results in at least a 21.74 % reduction in the <em>RMSE</em>. The method not only provides valuable feedback on machining errors of complex thin-walled parts but also offers technical support for the subsequent optimization and adjustment of machining strategies.</div></div>","PeriodicalId":16227,"journal":{"name":"Journal of Manufacturing Systems","volume":"81 ","pages":"Pages 208-221"},"PeriodicalIF":12.2,"publicationDate":"2025-05-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144146700","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}