Journal of Manufacturing Systems最新文献

{"title":"Knowledge-driven innovation in industrial maintenance: A neural-enhanced model-based definition framework for lifecycle maintenance process information propagation","authors":"Qidi Zhou ,&nbsp;Dong Zhou ,&nbsp;Chao Dai ,&nbsp;Jiayu Chen ,&nbsp;Ziyue Guo","doi":"10.1016/j.jmsy.2025.08.001","DOIUrl":"10.1016/j.jmsy.2025.08.001","url":null,"abstract":"<div><div>Under intensifying global competitive pressures, the digital strategic transformation of enterprises requires industrial information propagation across heterogeneous systems and lifecycle stages. These disparate transmission carriers and heterogeneous implementation mechanisms result in the inconsistent propagation of maintenance process information (MPI) in industrial information flows. These challenges render the structured data and knowledge in MPI, including maintenance activities, resource allocations, procedural instructions, and operational parameters, prone to ineffective dissemination across lifecycle phases and introduce risks of catastrophic operational failure. However, the direct application of current industrial information propagation methods, such as model-based definition (MBD) and intelligent information generation, encounters two obstacles: an incomplete standardization system for MPI definitions and construction and a mismatch between heterogeneous semistructured maintenance texts and the MPI. Therefore, a knowledge-driven neural-enhanced MBD framework for lifecycle MPI propagation is proposed. First, a lifecycle MPI propagation architecture is established to provide subsequent normative guidance. Second, an ontology-driven definition and construction method for MBD-based MPI is specified to address the obstacles posed by incomplete standardization systems. Third, an intelligent generation method for MBD-based MPI is constructed to overcome the obstacles of semantic mismatches. Finally, using aviation equipment as an example, the accuracy of the generated MPI and the feasibility of the innovative framework are verified via comparisons with current neural-enhanced models and results from multiple participants. The framework addresses lifecycle MPI propagation challenges through systematic knowledge formalization and neural-enhanced generation, advancing Industry 5.0’s vision of human-centric, resilient maintenance systems.</div></div>","PeriodicalId":16227,"journal":{"name":"Journal of Manufacturing Systems","volume":"82 ","pages":"Pages 976-999"},"PeriodicalIF":14.2,"publicationDate":"2025-08-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144831264","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":"DeepMS: A data-driven approach to machining process sequencing using transformers","authors":"Jaime Maqueda,&nbsp;David W. Rosen,&nbsp;Shreyes N. Melkote","doi":"10.1016/j.jmsy.2025.07.022","DOIUrl":"10.1016/j.jmsy.2025.07.022","url":null,"abstract":"<div><div>Efficient and intelligent machining process sequencing remains a key challenge in computer-aided process planning (CAPP). Traditional methods often rely on manually defined rules and explicit feature recognition, limiting their adaptability across diverse parts and evolving manufacturing environments. Recent advances in deep learning (DL), particularly in transformer-based sequence modeling, offer a promising alternative by enabling systems to learn sequencing logic directly from data without explicitly modeling complex rules. This paper presents a novel DL framework that predicts machining sequences directly from the 3D geometry of final parts. Operating on voxelized representations, the model generates an ordered sequence of machining operations, each associated with a volumetric shape representing the material removed from raw stock—eliminating the need for predefined features or rule-based logic. The framework integrates a transformer-based sequence autoencoder to model operation order and an encoder based on 3D convolutional neural networks (CNN) to map final part geometry to sequence representations. To efficiently handle high-dimensional voxelized data, a 3D CNN autoencoder is employed to compress voxelized removal volumes. Components of these pretrained models are combined into an inference pipeline that generates machining sequences directly from the final part geometry. Trained on a synthetic dataset of 1.08 million prismatic parts with embedded geometric precedence rules, the framework achieves a sequence prediction accuracy of 99.48 % and reconstructs final part geometry with a volumetric intersection-over-union (IoU) of 97.33 %. Results show the framework can generalize sequencing logic and material removal volumes from geometry data alone, offering a flexible and scalable approach to process planning and laying the foundation for future extensions in real-world manufacturing scenarios.</div></div>","PeriodicalId":16227,"journal":{"name":"Journal of Manufacturing Systems","volume":"82 ","pages":"Pages 947-963"},"PeriodicalIF":14.2,"publicationDate":"2025-08-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144809640","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":"Automated algorithm selection for Predictive Maintenance: Advances and challenges","authors":"Hendrik Engbers ,&nbsp;Michael Freitag","doi":"10.1016/j.jmsy.2025.06.023","DOIUrl":"10.1016/j.jmsy.2025.06.023","url":null,"abstract":"<div><div>Applications of Predictive Maintenance (PdM) in manufacturing systems with changing operating conditions still face significant challenges. In particular, the selection and application-specific configuration of prognostic algorithms often require expert knowledge and substantial computational resources, limiting scalability and broad adoption. Automated Machine Learning (AutoML) and Meta-Learning offer promising strategies to address these barriers; however, existing approaches frequently remain misaligned with the practical requirements of PdM in real-world industrial environments. This paper presents a systematic literature review of Meta-Learning techniques in the context of PdM. We first analyze the typical model development pipeline and emphasize the need for increased automation. Furthermore, general challenges associated with implementing PdM in industrial settings are discussed. After formalizing the problem as a Combined Algorithm Selection and Hyperparameter Optimization (CASH) task, a detailed literature analysis is conducted. The core contribution of this work is a structured assessment of Meta-Learning methods applied to time series forecasting and anomaly detection–two fundamental tasks in PdM. The review demonstrates the potential of Meta-Learning to improve algorithm and hyperparameter selection in PdM scenarios, while simultaneously identifying critical research gaps: (i) the underutilization of unsupervised approaches in low-label environments, (ii) the absence of adaptive methods capable of addressing dynamic industrial conditions, and (iii) the lack of robust integration strategies for deployment in operational settings. These findings provide a roadmap for future research at the intersection of Meta-Learning and industrial PdM.</div></div>","PeriodicalId":16227,"journal":{"name":"Journal of Manufacturing Systems","volume":"82 ","pages":"Pages 964-975"},"PeriodicalIF":14.2,"publicationDate":"2025-08-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144810204","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":"Superstructure-based optimization of robotic assembly lines","authors":"Gregor Wrobel ,&nbsp;Kerem Doğan ,&nbsp;Simon Hagemann ,&nbsp;Joshua Nelles ,&nbsp;Robert Scheffler","doi":"10.1016/j.jmsy.2025.06.021","DOIUrl":"10.1016/j.jmsy.2025.06.021","url":null,"abstract":"<div><div>Approaches to solve the assembly line balancing problem and its variations have been examined in research for several decades. Hybrid problems have replaced individual problems as the focus of consideration. Nevertheless, most of the theoretical models have not been applied in industry yet. This paper aims to close the research gap regarding the application of optimization strategies to production planning for fully automated robotic assembly lines. Therefore, we present a solution for solving a real-world hybrid problem that searches for the most cost-effective system. A balancing problem is solved for systems in which several robots can work together in parallel on fixtures or stationary joining units. The practical application includes the fact that the sequence of tasks is not given in a precedence graph, but instead, there is a sequence of processes that are implemented by several tasks. The planning of the task execution, optionally using different resources and in parallel, is part of the scheduling and equipment selection problem. We present a general domain model and a superstructure for robotic assembly lines and, based on this, an MILP formulation for the hybrid problem. Solutions for different cycle times were calculated for a real-world example and the results of this optimization are discussed and evaluated.</div></div>","PeriodicalId":16227,"journal":{"name":"Journal of Manufacturing Systems","volume":"82 ","pages":"Pages 786-808"},"PeriodicalIF":14.2,"publicationDate":"2025-08-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144772385","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":"Human-robot collaborative disassembly task planning for retired power battery based on Stackelberg game and multi-agent deep reinforcement learning","authors":"Xugang Zhang ,&nbsp;Chuang Liu ,&nbsp;Yong Yue ,&nbsp;Qingshan Gong ,&nbsp;Feng Ma ,&nbsp;Yan Wang","doi":"10.1016/j.jmsy.2025.07.024","DOIUrl":"10.1016/j.jmsy.2025.07.024","url":null,"abstract":"<div><div>With the increasing adoption of electric vehicles (EVs), the volume of retired power batteries has surged accordingly. In the context of sustainable development and a circular economy, the disassembly process of retired power battery for reuse is regarded as a crucial approach to addressing resource shortages and environmental pollution. To achieve the efficient disassembly of retired power batteries in human-robot collaborative (HRC) scenarios. Firstly, a mapping network between task units and task executors is established using a multilayer perceptron (MLP) neural network, based on the complexity of the tasks and the state of workers. Secondly, a Stackelberg Double Deep Q-Network (SDDQN) algorithm is proposed by integrating the leader-follower characteristics of the Stackelberg model with the Deep Q-Network (DQN) algorithm to address the task planning problem in HRC disassembly. Finally, the effectiveness of the proposed method is validated through two case studies. Compared to Nash Q-learning, Independent Q-learning, and the conventional DDQN algorithm, it demonstrates superior performance in terms of task completion time and average cumulative rewards. Additionally, the proposed method exhibits strong robustness against unexpected environmental disturbances.</div></div>","PeriodicalId":16227,"journal":{"name":"Journal of Manufacturing Systems","volume":"82 ","pages":"Pages 841-857"},"PeriodicalIF":14.2,"publicationDate":"2025-08-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144772386","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":"Leveraging large language models in next generation intelligent manufacturing: Retrospect and prospect","authors":"Yunfei Ma ,&nbsp;Shuai Zheng ,&nbsp;Zheng Yang ,&nbsp;Pai Zheng ,&nbsp;Jiewu Leng ,&nbsp;Jun Hong","doi":"10.1016/j.jmsy.2025.07.019","DOIUrl":"10.1016/j.jmsy.2025.07.019","url":null,"abstract":"<div><div>Industry 5.0, as the guiding ideology of the new generation intelligent manufacturing, points the way for global industrial transformation. It emphasizes the collaborative cooperation between humans, machines and intelligent systems, and places humans at the core of the industrial production process, aiming to create a more flexible, personalized and sustainable production paradigm. Large language model, as an advanced natural language processing technology, has received attention from researchers related to Industry 5.0 due to its ease of use and powerful language processing capability. LLM is considered to be one of the key enabling technologies to drive the development of Industry 5.0 and has great application potential. After a rigorous review of existing approaches, we find there is few existing survey papers that focuses on how LLM will drive the development of Industry 5.0 applications. Therefore, this paper provides a comprehensive review of the application of LLM in the field of Industry 5.0. Firstly, we conduct a literature review to explore the current state of research related to Industry 5.0. Subsequently, we analyze LLM-based technologies, synergizing LLMs with Industry 5.0 enablers and the applications of LLM in various domains of intelligent manufacturing. Finally, we explore the challenges of LLM in real-world scenarios and future research directions in the context of Industry 5.0. It is hoped that this study will contribute to the further development of LLM-based solutions in the context of Industry 5.0 and unite various efforts to achieve the vision of Industry 5.0.</div></div>","PeriodicalId":16227,"journal":{"name":"Journal of Manufacturing Systems","volume":"82 ","pages":"Pages 809-840"},"PeriodicalIF":14.2,"publicationDate":"2025-08-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144772387","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":"DAE-SWnet: Unsupervised internal defect segmentation through infrared thermography with scarce samples","authors":"Hao Shi ,&nbsp;Yifeng Pan ,&nbsp;Ruoxiang Gao ,&nbsp;Zhengchuan Guo ,&nbsp;Chengqian Zhang ,&nbsp;Peng Zhao","doi":"10.1016/j.jmsy.2025.07.016","DOIUrl":"10.1016/j.jmsy.2025.07.016","url":null,"abstract":"<div><div>Accurate defect detection is essential for manufacturing reliability, yet identifying internal defects remains challenging. While infrared thermography offers distinct advantages for internal inspection, its effectiveness is hindered by noise and uneven heating. Traditional image processing algorithms struggle with these nonlinearities, while supervised deep learning methods require large annotated datasets, usually impractical in industrial settings. To overcome data scarcity, we propose a novel infrared data amplification strategy leveraging the dynamic temperature evolution. By varying defect depth and heating duration, we generate 16000 thermal images from merely two defective samples using pulsed thermography. Furthermore, we introduce an unsupervised defect segmentation framework, Deep Autoencoder with Swin Transformer Wnet (DAE-SWnet). First, a deep autoencoder denoises thermal images, during which we discover a non-monotonic relationship between reconstruction loss and denoising performance. Next, Swin Transformer and Wnet are cohesively integrated with optimized encoder-decoder channels, to extract latent defect features from denoised images. These latent representations are postprocessed to obtain final segmentation results. Trained solely on artificially designed defects, our model exhibits exceptional performance across samples with varying materials and defect shapes. Moreover, comparative experiments demonstrate that the model achieves higher precision, stronger stability, and better generalization to real-world manufacturing processes. Specifically, it achieves 74.0 % IoU, 84.3 % F1-score, and 97.7 % accuracy, with an average inference time of 0.278 s, highlighting its superiority and practical potential in industrial scenarios where defective products are difficult to obtain and annotate.</div></div>","PeriodicalId":16227,"journal":{"name":"Journal of Manufacturing Systems","volume":"82 ","pages":"Pages 766-785"},"PeriodicalIF":14.2,"publicationDate":"2025-08-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144763807","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":"Tool condition monitoring and remaining useful life prediction based on multi-modal fusion and transfer learning under variable working conditions","authors":"Jiaqi Zhou ,&nbsp;Caixu Yue ,&nbsp;Wei Xia ,&nbsp;Xianli Liu ,&nbsp;Yanchang Zhou ,&nbsp;Zifeng Li ,&nbsp;Lihui Wang ,&nbsp;Steven Y. Liang","doi":"10.1016/j.jmsy.2025.07.021","DOIUrl":"10.1016/j.jmsy.2025.07.021","url":null,"abstract":"<div><div>Tool remaining useful life (RUL) prediction under various machining conditions constitutes crucial technology in the enhancement of machining quality and production efficiency. With the rapid development of intelligent manufacturing, the RUL prediction approach based on deep learning has been extensively employed due to its high efficacy and precision. Nevertheless, within the existing research, the input of single-modal data presents difficulties in comprehensively representing the tool wear feature information, and the generalization capacity of the model under variable working condition scenarios is limited, thereby constraining the practical application efficacy. The objective of this research is to propose a tool RUL prediction method based on multi-modal fusion transfer learning network with channel adaptive stochastic normalization (MFTLNCASN) to solve the existing problems. In the proposed method, the long short-term memory network (LSTM) is employed to extract the time series features of vibration signals and cutting force signals, thereby accomplishing multi-modal fusion within the feature level. A dual-channel prediction model is established by integrating star network (StarNet) and LSTM. Features are extracted from the fused signals and the surface texture images of the workpiece and then fused at the decision level. The channel adaptive stochastic normalization (CASN) method is devised to dynamically adjust the feature channel normalization strategy, to enhance the generalization ability of the model. Simultaneously, the fine-tuning technique is applied to reduce the disparity between the source domain and the target domain, facilitating high-precision RUL prediction under variable working conditions. Experiments were conducted using a face milling cutter. The effectiveness of the proposed method is verified under both constant and variable working conditions. The experimental outcomes demonstrate that MFTLNCASN exhibits superiority over the existing methods with respect to prediction accuracy and robustness. This research provides a new solution within the domain of tool condition monitoring and has significant practical guiding implications for the enhancement of machining quality and efficiency.</div></div>","PeriodicalId":16227,"journal":{"name":"Journal of Manufacturing Systems","volume":"82 ","pages":"Pages 730-747"},"PeriodicalIF":14.2,"publicationDate":"2025-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144749501","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":"Optimizing energy efficiency and productivity in industrial manufacturing: A simulation-based optimization approach with knowledge discovery","authors":"Thomas Schmitt ,&nbsp;Sergi Olives Juan ,&nbsp;Kaveh Amouzgar ,&nbsp;Lars Hanson ,&nbsp;Matías Urenda Moris","doi":"10.1016/j.jmsy.2025.07.008","DOIUrl":"10.1016/j.jmsy.2025.07.008","url":null,"abstract":"<div><div>Rising energy costs, energy supply uncertainties, and the sustainability crisis have intensified the need for energy efficiency in industrial manufacturing. This adds complexity to balancing traditional production goals such as productivity, quality, and cost. While prior studies address energy-intensive processes or throughput bottlenecks, they often lack integrated decision-support for evaluating optimal trade-offs. To address this gap, this study proposes a novel simulation-based multi-objective optimization framework combined with a knowledge discovery module, demonstrated in an industrial case study. The framework systematically identifies energy and productivity losses, evaluates improvement strategies to determine optimal trade-off solutions, and extracts actionable rules to guide decision making. Case study results show a 23.9% reduction in specific energy consumption and a 27.9% increase in throughput, while emphasizing the need to balance inventory levels. The approach offers a robust, data-driven method for supporting energy-efficient manufacturing. Future research will explore integration with real-time monitoring and extension to additional objectives such as costs and emissions.</div></div>","PeriodicalId":16227,"journal":{"name":"Journal of Manufacturing Systems","volume":"82 ","pages":"Pages 748-765"},"PeriodicalIF":14.2,"publicationDate":"2025-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144749500","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":"Multi-fidelity digital twin based optimization for aircraft overhaul shop scheduling","authors":"Mengnan Liu ,&nbsp;Shuiliang Fang ,&nbsp;Huiyue Dong","doi":"10.1016/j.jmsy.2025.07.018","DOIUrl":"10.1016/j.jmsy.2025.07.018","url":null,"abstract":"<div><div>Aircraft overhaul is of paramount importance for ensuring safety and reliability of aircraft throughout their entire lifecycle. The considerable number of overhaul tasks and prevalence of manual operations contribute to the elevated complexity and stochasticity of the aircraft overhaul shop scheduling problem (AOSSP), which is seldom considered in current researches. Digital twin (DT) has been proved to be an effective technical measure to simulate, evaluate, and predict the entire lifecycle of its physical entity in the fields of aerospace, automotive, infrastructure, etc. In most situations, the advantages of DT rely on the exact high-fidelity modeling of the physical systems to describe their features, behaviors, rules, etc. However, to optimize a large scale system as the aircraft overhaul shop, the high-fidelity digital twin model will be extremely computationally expensive. To this end, this paper proposes a multi-fidelity digital twin based optimization (MFDTBO) framework to solve AOSSP, which exploits the advantage of digital twin with acceptable level of computation cost. Firstly, the AOSSP is formulated mathematically after analyzing the aircraft overhaul process. Then the framework of MFDTBO is proposed, which embeds an improved hybrid genetic TABU search algorithm into multi-fidelity optimization with ordinal transformation and optimal sampling (MO²TOS). The AOSSP is solved in four stages, i.e. task assignment optimization with low fidelity digital twin, AOSSP optimization with low fidelity digital twin, AOSSP optimization with high fidelity digital twin, ultra-high fidelity digital twin simulation. The effectiveness of the proposed MFDTBO is verified and compared in different scales of test instances with different computation cost. A case study of applying the MFDTBO in aircraft overhaul digital twin system is provided to demonstrate the feasibility.</div></div>","PeriodicalId":16227,"journal":{"name":"Journal of Manufacturing Systems","volume":"82 ","pages":"Pages 700-729"},"PeriodicalIF":14.2,"publicationDate":"2025-07-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144724727","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}