Journal of Manufacturing Systems最新文献

An improved large language model and knowledge graph integration method for automated machining process base construction 一种改进的大语言模型与知识图集成的自动化加工过程库构建方法

IF 14.2 1区工程技术 Q1 ENGINEERING, INDUSTRIAL

Journal of Manufacturing Systems

Pub Date : 2026-01-28 DOI: 10.1016/j.jmsy.2026.01.016

Fu Yan , Jie Liu , Liang Guo , Li Liu , XiangYu Geng

The Machining Process Knowledge Base (MPKB) is foundational to intelligent process decision-making, directly impacting manufacturing efficiency and quality. While Large Language Models (LLMs) have shown promise in automated MPKB construction, they face a critical challenge in manufacturing: industrial knowledge graph (KG) schemas often exceed the context windows of lightweight LLMs deployable by small and medium-sized enterprises (SMEs). This limitation forces the construction process to operate with incomplete schema information, leading to missed entity relationships, semantic heterogeneity, and conceptual ambiguities in the MPKB. This study proposes an improved LLM-KG collaborative framework that overcomes these limitations through: (1) employing a staged, schema-decoupled extraction strategy, which enables open triple collection without injecting the full schema; (2) introducing a Code-Style knowledge representation method that efficiently encodes complex machining schemas, reducing the semantic load while maintaining structural integrity; and (3) constructing a retrieval-driven pipeline for semantic standardization that integrates dynamic schema segmentation and bidirectional validation, utilizing LLMs to achieve interpretable synonym merging and eliminate heterogeneity. This study empirically validated the proposed approach using machining process data provided by an aviation enterprise. Experimental results demonstrate that our framework achieves at least a 3.3% improvement in MPKB construction quality and a 25% increase in machining process quality metrics compared to the other baseline models. The implementation and data have been made available on GitHub to facilitate reproducibility and further research.

加工过程知识库（MPKB）是智能工艺决策的基础，直接影响到制造效率和质量。虽然大型语言模型（llm）在自动化MPKB构建中表现出了希望，但它们在制造业中面临着一个关键挑战：工业知识图（KG）模式通常超过了中小型企业（sme）可部署的轻量级llm的上下文窗口。这种限制迫使构造过程使用不完整的模式信息进行操作，从而导致MPKB中的实体关系缺失、语义异构和概念模糊。本研究提出了一种改进的LLM-KG协作框架，通过以下方式克服了这些限制：(1)采用分阶段、模式解耦的提取策略，在不注入完整模式的情况下实现开放的三重收集；(2)引入编码式知识表示方法，对复杂的加工模式进行高效编码，在保持结构完整性的同时减少语义负荷；(3)构建检索驱动的语义标准化管道，集成动态模式分割和双向验证，利用llm实现可解释同义词合并，消除异构性。利用某航空企业的加工工艺数据对该方法进行了实证验证。实验结果表明，与其他基准模型相比，我们的框架在MPKB构造质量方面至少提高了3.3%，在加工过程质量指标方面提高了25%。实现和数据已在GitHub上提供，以促进可重复性和进一步的研究。

{"title":"An improved large language model and knowledge graph integration method for automated machining process base construction","authors":"Fu Yan , Jie Liu , Liang Guo , Li Liu , XiangYu Geng","doi":"10.1016/j.jmsy.2026.01.016","DOIUrl":"10.1016/j.jmsy.2026.01.016","url":null,"abstract":"<div><div>The Machining Process Knowledge Base (MPKB) is foundational to intelligent process decision-making, directly impacting manufacturing efficiency and quality. While Large Language Models (LLMs) have shown promise in automated MPKB construction, they face a critical challenge in manufacturing: industrial knowledge graph (KG) schemas often exceed the context windows of lightweight LLMs deployable by small and medium-sized enterprises (SMEs). This limitation forces the construction process to operate with incomplete schema information, leading to missed entity relationships, semantic heterogeneity, and conceptual ambiguities in the MPKB. This study proposes an improved LLM-KG collaborative framework that overcomes these limitations through: (1) employing a staged, schema-decoupled extraction strategy, which enables open triple collection without injecting the full schema; (2) introducing a Code-Style knowledge representation method that efficiently encodes complex machining schemas, reducing the semantic load while maintaining structural integrity; and (3) constructing a retrieval-driven pipeline for semantic standardization that integrates dynamic schema segmentation and bidirectional validation, utilizing LLMs to achieve interpretable synonym merging and eliminate heterogeneity. This study empirically validated the proposed approach using machining process data provided by an aviation enterprise. Experimental results demonstrate that our framework achieves at least a 3.3% improvement in MPKB construction quality and a 25% increase in machining process quality metrics compared to the other baseline models. The implementation and data have been made available on GitHub to facilitate reproducibility and further research.</div></div>","PeriodicalId":16227,"journal":{"name":"Journal of Manufacturing Systems","volume":"85 ","pages":"Pages 318-337"},"PeriodicalIF":14.2,"publicationDate":"2026-01-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146078235","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 physics-informed embodied intelligence framework for smart garment manufacturing 智能服装制造的物理信息嵌入智能框架

IF 14.2 1区工程技术 Q1 ENGINEERING, INDUSTRIAL

Journal of Manufacturing Systems

Pub Date : 2026-01-27 DOI: 10.1016/j.jmsy.2026.01.015

Shufei Li , Ao Dong , Ray Wai Man Kong , Cheng Liu

Embodied intelligence systems are becoming integral to smart manufacturing, driven by the human-centric, sustainable, and resilient vision of Industry 5.0. While robots have achieved remarkable success in manipulating rigid objects, robust handling of deformable fabrics in garment manufacturing remains challenging, particularly for pick-and-place operations required in sewing processes. In production environments, robots often struggle to adapt gripping strategies to different fabric properties and deformation behavior, limiting smart manufacturing development. To address these challenges, this paper proposes a physics-informed embodied intelligence framework tailored for smart garment manufacturing. The system integrates task-aware planning, enabled by vision–language understanding of the surrounding environment, with physics-informed manipulation, which incorporates material property inference and finite element analysis. This integration enables robots to perform garment manufacturing tasks with greater adaptability, precision, and intelligence. Experiments conducted in real-world scenarios validate the effectiveness of the proposed framework in supporting flexible and intelligent production processes.

在以人为中心、可持续和弹性的工业5.0愿景的推动下，具身智能系统正在成为智能制造不可或缺的一部分。虽然机器人在操纵刚性物体方面取得了显著的成功，但在服装制造中对可变形织物的稳健处理仍然具有挑战性，特别是在缝制过程中需要的取放操作。在生产环境中，机器人往往难以适应不同织物性能和变形行为的抓取策略，限制了智能制造的发展。为了解决这些挑战，本文提出了一个为智能服装制造量身定制的物理知情的具身智能框架。该系统集成了任务感知规划，通过对周围环境的视觉语言理解，以及物理信息操作，其中包括材料特性推断和有限元分析。这种集成使机器人能够以更高的适应性、精度和智能执行服装制造任务。在实际场景中进行的实验验证了所提出的框架在支持灵活和智能生产过程方面的有效性。

{"title":"A physics-informed embodied intelligence framework for smart garment manufacturing","authors":"Shufei Li , Ao Dong , Ray Wai Man Kong , Cheng Liu","doi":"10.1016/j.jmsy.2026.01.015","DOIUrl":"10.1016/j.jmsy.2026.01.015","url":null,"abstract":"<div><div>Embodied intelligence systems are becoming integral to smart manufacturing, driven by the human-centric, sustainable, and resilient vision of Industry 5.0. While robots have achieved remarkable success in manipulating rigid objects, robust handling of deformable fabrics in garment manufacturing remains challenging, particularly for pick-and-place operations required in sewing processes. In production environments, robots often struggle to adapt gripping strategies to different fabric properties and deformation behavior, limiting smart manufacturing development. To address these challenges, this paper proposes a physics-informed embodied intelligence framework tailored for smart garment manufacturing. The system integrates task-aware planning, enabled by vision–language understanding of the surrounding environment, with physics-informed manipulation, which incorporates material property inference and finite element analysis. This integration enables robots to perform garment manufacturing tasks with greater adaptability, precision, and intelligence. Experiments conducted in real-world scenarios validate the effectiveness of the proposed framework in supporting flexible and intelligent production processes.</div></div>","PeriodicalId":16227,"journal":{"name":"Journal of Manufacturing Systems","volume":"85 ","pages":"Pages 307-317"},"PeriodicalIF":14.2,"publicationDate":"2026-01-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146078236","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

Automated extraction of comprehensive digital twin models for smart manufacturing systems 智能制造系统综合数字孪生模型的自动提取

IF 14.2 1区工程技术 Q1 ENGINEERING, INDUSTRIAL

Journal of Manufacturing Systems

Pub Date : 2026-01-26 DOI: 10.1016/j.jmsy.2026.01.012

Atieh Khodadadi , Sanja Lazarova-Molnar

Manufacturing systems involve multiple, often conflicting, objectives referred to as performance indicators, including production efficiency, resource utilization, energy consumption, carbon emissions, and waste reduction, which correspond to different dimensions of the system, such as time, energy consumption, and waste generation aspects. Digital Twins have emerged as a powerful tool, integrating data-driven simulation and analysis of complex systems, such as manufacturing systems. Process Mining (PM), along with data analysis, enables the automatic discovery of executable discrete-event simulation models directly from production event logs. These data-driven models are the key to enabling near-real-time Digital Twins of discrete-event systems. Stochastic Petri Nets (SPNs) offer a robust and intuitive modeling formalism well-suited for representing the extracted models derived from PM, particularly in the context of manufacturing systems. However, standard SPNs face challenges in incorporating dimensions beyond time, such as energy consumption and waste generation. This limitation often results in suboptimal decision-making and reduced system efficiency. In this paper, we propose a Comprehensive Digital Twin (CDT) framework that employs Multi-Flow Process Mining (MFPM) to automatically extract Multidimensional Stochastic Petri Nets (MDSPNs) as underlying models of manufacturing systems. To support the modeling and simulation of extracted MDSPNs, we introduce and utilize our tool, MDPySPN. The CDT framework supports multi-objective decision-making for various performance indicators of the system. Through an illustrative case study of hot forging process chains, we showcase the development of CDT for time, energy consumption, and waste generation dimensions. We further illustrate the utilization of CDT to analyze and support decision-making to enhance the case study system according to its objectives.

制造系统涉及多个通常相互冲突的目标，这些目标被称为绩效指标，包括生产效率、资源利用、能源消耗、碳排放和废物减少，这些目标对应于系统的不同维度，例如时间、能源消耗和废物产生方面。数字孪生已经成为一种强大的工具，集成了数据驱动的模拟和复杂系统的分析，如制造系统。流程挖掘（Process Mining， PM）与数据分析一起，可以直接从生产事件日志中自动发现可执行的离散事件模拟模型。这些数据驱动模型是实现离散事件系统的近实时数字孪生的关键。随机Petri网（spn）提供了一种鲁棒和直观的建模形式，非常适合于表示从PM中提取的模型，特别是在制造系统的背景下。然而，标准spn在纳入时间以外的维度方面面临挑战，例如能源消耗和废物产生。这种限制通常会导致次优决策和降低系统效率。在本文中，我们提出了一个综合数字孪生（CDT）框架，该框架采用多流程挖掘（MFPM）来自动提取多维随机Petri网（mdspn）作为制造系统的底层模型。为了支持提取的mdspn的建模和仿真，我们引入并利用了我们的工具MDPySPN。CDT框架支持系统各种绩效指标的多目标决策。通过热锻工艺链的一个说明性案例研究，我们展示了CDT在时间、能源消耗和废物产生方面的发展。我们进一步说明利用CDT来分析和支持决策，以根据其目标增强案例研究系统。

{"title":"Automated extraction of comprehensive digital twin models for smart manufacturing systems","authors":"Atieh Khodadadi , Sanja Lazarova-Molnar","doi":"10.1016/j.jmsy.2026.01.012","DOIUrl":"10.1016/j.jmsy.2026.01.012","url":null,"abstract":"<div><div>Manufacturing systems involve multiple, often conflicting, objectives referred to as performance indicators, including production efficiency, resource utilization, energy consumption, carbon emissions, and waste reduction, which correspond to different dimensions of the system, such as time, energy consumption, and waste generation aspects. Digital Twins have emerged as a powerful tool, integrating data-driven simulation and analysis of complex systems, such as manufacturing systems. Process Mining (PM), along with data analysis, enables the automatic discovery of executable discrete-event simulation models directly from production event logs. These data-driven models are the key to enabling near-real-time Digital Twins of discrete-event systems. Stochastic Petri Nets (SPNs) offer a robust and intuitive modeling formalism well-suited for representing the extracted models derived from PM, particularly in the context of manufacturing systems. However, standard SPNs face challenges in incorporating dimensions beyond time, such as energy consumption and waste generation. This limitation often results in suboptimal decision-making and reduced system efficiency. In this paper, we propose a Comprehensive Digital Twin (CDT) framework that employs Multi-Flow Process Mining (MFPM) to automatically extract Multidimensional Stochastic Petri Nets (MDSPNs) as underlying models of manufacturing systems. To support the modeling and simulation of extracted MDSPNs, we introduce and utilize our tool, MDPySPN. The CDT framework supports multi-objective decision-making for various performance indicators of the system. Through an illustrative case study of hot forging process chains, we showcase the development of CDT for time, energy consumption, and waste generation dimensions. We further illustrate the utilization of CDT to analyze and support decision-making to enhance the case study system according to its objectives.</div></div>","PeriodicalId":16227,"journal":{"name":"Journal of Manufacturing Systems","volume":"85 ","pages":"Pages 287-306"},"PeriodicalIF":14.2,"publicationDate":"2026-01-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146078234","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

Wafer defect semantic reasoning via a three-stage retrieval-augmented system in semiconductor manufacturing 基于三阶段检索增强系统的半导体制造晶圆缺陷语义推理

IF 14.2 1区工程技术 Q1 ENGINEERING, INDUSTRIAL

Journal of Manufacturing Systems

Pub Date : 2026-01-23 DOI: 10.1016/j.jmsy.2025.12.026

Xinting Liao , Jie Zhang , Youlong Lyu , Junliang Wang

Wafer defect detection is a crucial part of semiconductor manufacturing, requiring both high generalization across distribution shifts and interpretable explanations for diverse engineering roles. Existing deep learning methods often fall short in semantic reasoning and domain transferability. To address these challenges, we propose WaferDSR-RAG, a novel retrieval-augmented framework that integrates a bimodal wafer defect knowledge graph (BWDKG) with a three-stage semantic adaptation strategy, comprising defect visual-semantic alignment, defect-relevant knowledge retrieval and screening, and role-adaptive explanation generation. This design allows the system to dynamically adapt to unseen defect distributions and generate expert-aligned explanations tailored to different engineering responsibilities. Extensive experiments on both in-distribution and cross-fab wafer maps demonstrate that WaferDSR-RAG consistently outperforms state-of-the-art baselines in both detection accuracy and explanation quality. Compared to GPT-4o and Gemini 1.5 Pro, our method generates more semantically accurate and practically useful explanations for engineers with different roles, as validated by automatic metrics and domain expert evaluations. WaferDSR-RAG represents a scalable solution for wafer defect detection and reasoning in real-world semiconductor production.

晶圆缺陷检测是半导体制造的重要组成部分，需要在分布转移和不同工程角色的可解释的解释之间具有高度的通用性。现有的深度学习方法在语义推理和领域可转移性方面存在不足。为了解决这些挑战，我们提出了一种新的检索增强框架WaferDSR-RAG，该框架将双峰晶圆缺陷知识图（BWDKG）与三阶段语义自适应策略集成在一起，包括缺陷视觉语义对齐、缺陷相关知识检索和筛选以及角色自适应解释生成。这种设计允许系统动态地适应不可见的缺陷分布，并生成针对不同工程责任的专家一致的解释。在分布和跨晶圆图上进行的大量实验表明，WaferDSR-RAG在检测精度和解释质量方面始终优于最先进的基线。与gpt - 40和Gemini 1.5 Pro相比，我们的方法为不同角色的工程师生成了更准确的语义和实用的解释，并得到了自动度量和领域专家评估的验证。WaferDSR-RAG为实际半导体生产中的晶圆缺陷检测和推理提供了可扩展的解决方案。

{"title":"Wafer defect semantic reasoning via a three-stage retrieval-augmented system in semiconductor manufacturing","authors":"Xinting Liao , Jie Zhang , Youlong Lyu , Junliang Wang","doi":"10.1016/j.jmsy.2025.12.026","DOIUrl":"10.1016/j.jmsy.2025.12.026","url":null,"abstract":"<div><div>Wafer defect detection is a crucial part of semiconductor manufacturing, requiring both high generalization across distribution shifts and interpretable explanations for diverse engineering roles. Existing deep learning methods often fall short in semantic reasoning and domain transferability. To address these challenges, we propose WaferDSR-RAG, a novel retrieval-augmented framework that integrates a bimodal wafer defect knowledge graph (BWDKG) with a three-stage semantic adaptation strategy, comprising defect visual-semantic alignment, defect-relevant knowledge retrieval and screening, and role-adaptive explanation generation. This design allows the system to dynamically adapt to unseen defect distributions and generate expert-aligned explanations tailored to different engineering responsibilities. Extensive experiments on both in-distribution and cross-fab wafer maps demonstrate that WaferDSR-RAG consistently outperforms state-of-the-art baselines in both detection accuracy and explanation quality. Compared to GPT-4o and Gemini 1.5 Pro, our method generates more semantically accurate and practically useful explanations for engineers with different roles, as validated by automatic metrics and domain expert evaluations. WaferDSR-RAG represents a scalable solution for wafer defect detection and reasoning in real-world semiconductor production.</div></div>","PeriodicalId":16227,"journal":{"name":"Journal of Manufacturing Systems","volume":"85 ","pages":"Pages 269-286"},"PeriodicalIF":14.2,"publicationDate":"2026-01-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146034981","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

Corrigendum to “ChatCNC: Conversational machine monitoring via large language model and real-time data retrieval augmented generation” [J Manuf Syst 79 (2025) 504–514] “ChatCNC：通过大型语言模型和实时数据检索增强生成的会话机器监控”的勘误表[J] Manuf system 79 (2025) 504-514]

IF 14.2 1区工程技术 Q1 ENGINEERING, INDUSTRIAL

Journal of Manufacturing Systems

Pub Date : 2026-01-22 DOI: 10.1016/j.jmsy.2026.01.008

Jurim Jeon , Yuseop Sim , Hojun Lee , Changheon Han , Dongjun Yun , Eunseob Kim , Shreya Laxmi Nagendra , Martin B.G. Jun , Yangjin Kim , Sang Won Lee , Jiho Lee

引用次数: 0

Large language models in human-robot collaboration: A systematic review, trends, and challenges 人机协作中的大型语言模型：系统回顾、趋势和挑战

IF 14.2 1区工程技术 Q1 ENGINEERING, INDUSTRIAL

Journal of Manufacturing Systems

Pub Date : 2026-01-22 DOI: 10.1016/j.jmsy.2026.01.011

Gang Yuan , Xiaojun Liu , Maohua Xiao , Jinhua Xiao , Lihui Wang

The integration of large language models (LLMs) into human-robot collaboration (HRC) represents a paradigm shift toward cognitive manufacturing under Industry 5.0. This systematic review analyzes 1278 publications to elucidate the enabling mechanisms, application hierarchies, and persistent challenges of LLM-enhanced HRC. We identify three agent paradigms, each tailored to structured, interactive, and sensor-rich scenarios, respectively. Furthermore, we propose a layered training framework combining domain-adaptive pre-training and scenario-specific fine-tuning to bridge general LLMs capabilities with industrial demands. Our findings reveal that LLMs significantly advance HRC in environmental perception, long-horizon task planning, and embodied interaction, yet face critical challenges in dynamic adaptability, multimodal fusion, and value alignment. This study provides a structured reference for developing robust, interpretable, and human-centric collaborative systems, outlining five key research directions to guide future advancements in cognitive manufacturing.

将大型语言模型（llm）集成到人机协作（HRC）中代表了工业5.0下向认知制造的范式转变。本系统综述分析了1278篇出版物，以阐明llm增强的HRC的启用机制、应用层次和持续挑战。我们确定了三种智能体范式，分别适用于结构化、交互式和传感器丰富的场景。此外，我们提出了一个结合领域自适应预训练和场景特定微调的分层培训框架，以将llm的一般能力与工业需求联系起来。研究结果表明，llm在环境感知、长期任务规划和具身互动方面显著推进了HRC，但在动态适应性、多模态融合和价值一致性方面面临严峻挑战。本研究为开发稳健的、可解释的、以人为中心的协作系统提供了结构化的参考，概述了五个关键的研究方向，以指导认知制造的未来发展。

{"title":"Large language models in human-robot collaboration: A systematic review, trends, and challenges","authors":"Gang Yuan , Xiaojun Liu , Maohua Xiao , Jinhua Xiao , Lihui Wang","doi":"10.1016/j.jmsy.2026.01.011","DOIUrl":"10.1016/j.jmsy.2026.01.011","url":null,"abstract":"<div><div>The integration of large language models (LLMs) into human-robot collaboration (HRC) represents a paradigm shift toward cognitive manufacturing under Industry 5.0. This systematic review analyzes 1278 publications to elucidate the enabling mechanisms, application hierarchies, and persistent challenges of LLM-enhanced HRC. We identify three agent paradigms, each tailored to structured, interactive, and sensor-rich scenarios, respectively. Furthermore, we propose a layered training framework combining domain-adaptive pre-training and scenario-specific fine-tuning to bridge general LLMs capabilities with industrial demands. Our findings reveal that LLMs significantly advance HRC in environmental perception, long-horizon task planning, and embodied interaction, yet face critical challenges in dynamic adaptability, multimodal fusion, and value alignment. This study provides a structured reference for developing robust, interpretable, and human-centric collaborative systems, outlining five key research directions to guide future advancements in cognitive manufacturing.</div></div>","PeriodicalId":16227,"journal":{"name":"Journal of Manufacturing Systems","volume":"85 ","pages":"Pages 249-268"},"PeriodicalIF":14.2,"publicationDate":"2026-01-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146035033","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

The SHOP4CF modular reference architecture for flexible process-oriented, data-driven smart manufacturing 面向灵活流程、数据驱动的智能制造的SHOP4CF模块化参考架构

IF 14.2 1区工程技术 Q1 ENGINEERING, INDUSTRIAL

Journal of Manufacturing Systems

Pub Date : 2026-01-20 DOI: 10.1016/j.jmsy.2026.01.010

Paul Grefen , Michał Zimniewicz , Irene Vanderfeesten , Kostas Traganos , Pieter Becue , Anders Pedersen , Genessis Perez Rivera

Context

Organizations in the smart industry domain face an increasing complexity of their functions and processes, both in the intra- and inter-organizational scopes. This has a direct effect on the digital systems that support their operations: they grow more complex too. At the same time, the organizations need to increase their agility: they must be able to flexibly adapt their processes to market changes. Especially SMEs in the manufacturing domain get lost in this combination of complexity and changeability.

Objectives

To help SME organizations in the smart manufacturing domain with their digital transformation, we develop the SHOP4CF modular reference architecture for digital manufacturing support.

Methods

We develop the SHOP4CF base architecture in an iterative way by application and evaluation in 36 real-world industrial cases, organized in three waves. We base our design partly on successful existing work, specifically the outcomes of the HORSE EU project, and align it with main manufacturing standards like ISA-95 and RAMI 4.0. We next distill the SHOP4CF reference architecture by abstracting the SHOP4CF base architecture, based on explicit design principles. We then specialize the reference architecture for process-oriented and data-driven manufacturing.

Results

The result of our work is a modular, flexible software reference architecture for smart manufacturing solutions. To facilitate its use, the reference architecture is coupled with manufacturing software life cycle models. Centered on a component marketplace, the life cycle for functional module developers is linked to the life cycle for module users, including explicit attention to the role of technology integrators. To illustrate its applicability, we describe three application cases in this paper.

Conclusion

The reference architecture provides a demonstrated point of departure for SMEs in the manufacturing domain to design their digital support in a complex and dynamic industry ecosystem. The modularity of the architecture and its coupling to software life cycles provide a new level of flexibility.

智能工业领域的组织在组织内部和组织间都面临着越来越复杂的功能和流程。这对支持它们运作的数字系统产生了直接影响：它们也变得更加复杂。同时，组织需要增加他们的敏捷性：他们必须能够灵活地调整他们的过程以适应市场变化。尤其是制造业领域的中小企业，在这种复杂性和可变性的组合中迷失了方向。为了帮助智能制造领域的中小企业组织进行数字化转型，我们开发了用于数字化制造支持的SHOP4CF模块化参考架构。方法通过对36个实际工业案例的应用和评估，以迭代的方式开发SHOP4CF基础架构。我们的设计部分基于成功的现有工作，特别是HORSE欧盟项目的成果，并使其与ISA-95和RAMI 4.0等主要制造标准保持一致。接下来，我们根据显式设计原则，通过抽象SHOP4CF基本体系结构，提炼出SHOP4CF参考体系结构。然后，我们专门研究面向过程和数据驱动制造的参考体系结构。我们的工作成果是智能制造解决方案的模块化、灵活的软件参考架构。为了便于使用，参考体系结构与制造软件生命周期模型相结合。以组件市场为中心，功能模块开发人员的生命周期与模块用户的生命周期相关联，包括对技术集成商角色的明确关注。为了说明其适用性，本文描述了三个应用案例。该参考架构为制造业领域的中小企业在复杂而动态的行业生态系统中设计其数字支持提供了一个示范出发点。体系结构的模块化及其与软件生命周期的耦合提供了新的灵活性级别。

{"title":"The SHOP4CF modular reference architecture for flexible process-oriented, data-driven smart manufacturing","authors":"Paul Grefen , Michał Zimniewicz , Irene Vanderfeesten , Kostas Traganos , Pieter Becue , Anders Pedersen , Genessis Perez Rivera","doi":"10.1016/j.jmsy.2026.01.010","DOIUrl":"10.1016/j.jmsy.2026.01.010","url":null,"abstract":"<div><h3>Context</h3><div>Organizations in the smart industry domain face an increasing complexity of their functions and processes, both in the intra- and inter-organizational scopes. This has a direct effect on the digital systems that support their operations: they grow more complex too. At the same time, the organizations need to increase their agility: they must be able to flexibly adapt their processes to market changes. Especially SMEs in the manufacturing domain get lost in this combination of complexity and changeability.</div></div><div><h3>Objectives</h3><div>To help SME organizations in the smart manufacturing domain with their digital transformation, we develop the SHOP4CF modular reference architecture for digital manufacturing support.</div></div><div><h3>Methods</h3><div>We develop the SHOP4CF base architecture in an iterative way by application and evaluation in 36 real-world industrial cases, organized in three waves. We base our design partly on successful existing work, specifically the outcomes of the HORSE EU project, and align it with main manufacturing standards like ISA-95 and RAMI 4.0. We next distill the SHOP4CF reference architecture by abstracting the SHOP4CF base architecture, based on explicit design principles. We then specialize the reference architecture for process-oriented and data-driven manufacturing.</div></div><div><h3>Results</h3><div>The result of our work is a modular, flexible software reference architecture for smart manufacturing solutions. To facilitate its use, the reference architecture is coupled with manufacturing software life cycle models. Centered on a component marketplace, the life cycle for functional module developers is linked to the life cycle for module users, including explicit attention to the role of technology integrators. To illustrate its applicability, we describe three application cases in this paper.</div></div><div><h3>Conclusion</h3><div>The reference architecture provides a demonstrated point of departure for SMEs in the manufacturing domain to design their digital support in a complex and dynamic industry ecosystem. The modularity of the architecture and its coupling to software life cycles provide a new level of flexibility.</div></div>","PeriodicalId":16227,"journal":{"name":"Journal of Manufacturing Systems","volume":"85 ","pages":"Pages 227-247"},"PeriodicalIF":14.2,"publicationDate":"2026-01-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146035034","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

A4PS: Agentic AI-assisted advanced planning and scheduling with large language models for smart manufacturing A4PS：智能制造大语言模型，人工智能辅助高级规划调度

IF 14.2 1区工程技术 Q1 ENGINEERING, INDUSTRIAL

Journal of Manufacturing Systems

Pub Date : 2026-01-16 DOI: 10.1016/j.jmsy.2026.01.003

Mingxing Li , Qu Zhou , Wanshan Li , Ting Qu , Maolin Yang , Pingyu Jiang

Advanced Planning and Scheduling (APS) for manufacturing systems is becoming more complex against the backdrop of intelligent transformation and increasing demand for customisation. In real-world APS applications subject to multi-source dynamics, objective alterations, constraints removals/additions, algorithm upgrades are inevitable. Such structural changes of APS, requiring seamless coordination among experts such as production managers, modelling engineers, algorithm developers, are often lengthy and less flexible. This poses new challenges in cross-domain/inter-process coordination and rapid multi-disciplinary knowledge integration/reuse. This paper proposes a novel Agentic AI-Assisted APS (A⁴PS) framework, utilising Large Language Models (LLMs) and agents to assist modification/update processes of APS. Firstly, a multi-agentic AI-enabled workflow is designed following standard operating procedure of APS to facilitate the cross-domain agent coordination. Secondly, a multi-step knowledge augmentation method is proposed to endow LLM agents with specialised APS knowledge. Thirdly, a Retrieval-Augmented Generation (RAG) and Chain of Thought (CoT)-enhanced method is developed for knowledge use and interaction. Experiments are conducted with an APS dataset which is created based on classical APS cases and manufacturing researchers. Compared with basic LLMs, A⁴PS exhibited substantially superior performance across both basic and complex cases in metrics such as modelling task success rate, absolute percentage error of solution results, optimisation algorithm code logic completion rate, and code executability rate. Case study demonstrates that A⁴PS enables LLMs to coordinate, learn APS knowledge, and imitate experts in the reasoning process, and ultimately realise APS assistance using natural language. This work proposes a novel solution that uses LLMs and agentic AI to assist APS modification/update process, contributing to AI-driven smart manufacturing in Industry 4.0.

在智能转型和不断增长的定制需求的背景下，制造系统的高级计划和调度（APS）变得越来越复杂。在现实世界的APS应用中，受制于多源动态、目标改变、约束移除/添加、算法升级是不可避免的。APS的这种结构变化，需要生产经理、建模工程师、算法开发人员等专家之间的无缝协调，通常是冗长而不灵活的。这对跨领域/跨过程的协调和多学科知识的快速集成/重用提出了新的挑战。本文提出了一种新的代理人工智能辅助APS （A4PS）框架，利用大型语言模型（LLMs）和代理来辅助APS的修改/更新过程。首先，根据APS的标准操作流程，设计了一个支持ai的多代理工作流，以促进跨域代理的协调。其次，提出了一种多步知识增强方法，赋予LLM agent专门的APS知识。第三，提出了一种检索增强生成（RAG）和思维链（CoT）增强的知识使用和交互方法。基于经典APS案例和制造业研究人员创建的APS数据集进行了实验。与基本llm相比，A4PS在基本和复杂情况下都表现出明显更好的性能，例如建模任务成功率、解决结果的绝对错误率、优化算法代码逻辑完成率和代码可执行率。案例研究表明，A4PS使llm能够协调、学习APS知识，并在推理过程中模仿专家，最终实现使用自然语言的APS辅助。本研究提出了一种新颖的解决方案，使用llm和人工智能代理来协助APS修改/更新过程，为工业4.0中人工智能驱动的智能制造做出贡献。

{"title":"A4PS: Agentic AI-assisted advanced planning and scheduling with large language models for smart manufacturing","authors":"Mingxing Li , Qu Zhou , Wanshan Li , Ting Qu , Maolin Yang , Pingyu Jiang","doi":"10.1016/j.jmsy.2026.01.003","DOIUrl":"10.1016/j.jmsy.2026.01.003","url":null,"abstract":"<div><div>Advanced Planning and Scheduling (APS) for manufacturing systems is becoming more complex against the backdrop of intelligent transformation and increasing demand for customisation. In real-world APS applications subject to multi-source dynamics, objective alterations, constraints removals/additions, algorithm upgrades are inevitable. Such structural changes of APS, requiring seamless coordination among experts such as production managers, modelling engineers, algorithm developers, are often lengthy and less flexible. This poses new challenges in cross-domain/inter-process coordination and rapid multi-disciplinary knowledge integration/reuse. This paper proposes a novel <strong><em>A</em></strong><em>gentic</em> <strong><em>A</em></strong><em>I-</em><strong><em>A</em></strong><em>ssisted</em> <strong><em>APS</em></strong> (A<sup>4</sup>PS) framework, utilising Large Language Models (LLMs) and agents to assist modification/update processes of APS. Firstly, a multi-agentic AI-enabled workflow is designed following standard operating procedure of APS to facilitate the cross-domain agent coordination. Secondly, a multi-step knowledge augmentation method is proposed to endow LLM agents with specialised APS knowledge. Thirdly, a Retrieval-Augmented Generation (RAG) and Chain of Thought (CoT)-enhanced method is developed for knowledge use and interaction. Experiments are conducted with an APS dataset which is created based on classical APS cases and manufacturing researchers. Compared with basic LLMs, A<sup>4</sup>PS exhibited substantially superior performance across both basic and complex cases in metrics such as modelling task success rate, absolute percentage error of solution results, optimisation algorithm code logic completion rate, and code executability rate. Case study demonstrates that A<sup>4</sup>PS enables LLMs to coordinate, learn APS knowledge, and imitate experts in the reasoning process, and ultimately realise APS assistance using natural language. This work proposes a novel solution that uses LLMs and agentic AI to assist APS modification/update process, contributing to AI-driven smart manufacturing in Industry 4.0.</div></div>","PeriodicalId":16227,"journal":{"name":"Journal of Manufacturing Systems","volume":"85 ","pages":"Pages 207-226"},"PeriodicalIF":14.2,"publicationDate":"2026-01-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145978607","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

Computer-aided manufacturing in reality: augmented reality-enabled precision motion planning for robotic repair of complex shapes 现实中的计算机辅助制造：用于复杂形状的机器人修复的增强现实支持的精确运动规划

IF 14.2 1区工程技术 Q1 ENGINEERING, INDUSTRIAL

Journal of Manufacturing Systems

Pub Date : 2026-01-16 DOI: 10.1016/j.jmsy.2026.01.006

Chih-Hsing Chu , Cong-Yun Lai , Chih-Husan Shih

Augmented Reality (AR) is increasingly used as a user interface for robot motion planning in manufacturing environments. However, the accuracy of 3D pointing (waypoint input) in most AR applications is insufficient for tasks that demand high precision. This study introduces the concept of Computer-Aided Manufacturing in Reality (CAMiR) to enable high-precision motion planning and path computation for robotic welding and grinding operations. Human interactions within the AR scene, intended for visualization purposes, are separated from spatial pointing, used for defining robot movements. The points manually specified using a self-tracking tool are aligned with the measurements obtained from an external laser profiler through 3D registration, thereby reducing positional deviations relative to the workpiece. Geometric modeling functions are accessible through AR interfaces to construct geometric elements from the aligned points, which define processing regions and generate corresponding paths. A prototyping system implementing CAMiR allows operators to plan and simulate robot motion directly on real parts without the need to reconstruct CAD models. Evaluation results on test workpieces indicate that the pointing accuracy is significantly improved compared with previous studies and the ground surface exhibits superior geometric precision. The repair of complex 3D defects using robotic motions generated by the system verifies the effectiveness of the proposed approach and highlights its potential to enhance the practical applicability of AR interfaces in high-precision manufacturing tasks.

增强现实（AR）越来越多地被用作制造环境中机器人运动规划的用户界面。然而，在大多数AR应用中，3D指向（航路点输入）的精度不足以满足要求高精度的任务。本研究引入现实中计算机辅助制造（CAMiR）的概念，实现机器人焊接和磨削操作的高精度运动规划和路径计算。用于可视化目的的AR场景中的人类交互与用于定义机器人运动的空间指向分离。使用自跟踪工具手动指定的点与通过3D配准从外部激光轮廓仪获得的测量值对齐，从而减少相对于工件的位置偏差。通过AR接口访问几何建模功能，从对齐点构建几何元素，定义加工区域并生成相应路径。实现CAMiR的原型系统允许操作员直接在真实部件上规划和模拟机器人运动，而无需重建CAD模型。对测试工件的评价结果表明，与以往的研究相比，该方法的指向精度有了显著提高，地表几何精度也有所提高。利用系统产生的机器人运动修复复杂的3D缺陷验证了所提出方法的有效性，并突出了其在增强现实接口在高精度制造任务中的实际适用性的潜力。

{"title":"Computer-aided manufacturing in reality: augmented reality-enabled precision motion planning for robotic repair of complex shapes","authors":"Chih-Hsing Chu , Cong-Yun Lai , Chih-Husan Shih","doi":"10.1016/j.jmsy.2026.01.006","DOIUrl":"10.1016/j.jmsy.2026.01.006","url":null,"abstract":"<div><div>Augmented Reality (AR) is increasingly used as a user interface for robot motion planning in manufacturing environments. However, the accuracy of 3D pointing (waypoint input) in most AR applications is insufficient for tasks that demand high precision. This study introduces the concept of Computer-Aided Manufacturing in Reality (CAMiR) to enable high-precision motion planning and path computation for robotic welding and grinding operations. Human interactions within the AR scene, intended for visualization purposes, are separated from spatial pointing, used for defining robot movements. The points manually specified using a self-tracking tool are aligned with the measurements obtained from an external laser profiler through 3D registration, thereby reducing positional deviations relative to the workpiece. Geometric modeling functions are accessible through AR interfaces to construct geometric elements from the aligned points, which define processing regions and generate corresponding paths. A prototyping system implementing CAMiR allows operators to plan and simulate robot motion directly on real parts without the need to reconstruct CAD models. Evaluation results on test workpieces indicate that the pointing accuracy is significantly improved compared with previous studies and the ground surface exhibits superior geometric precision. The repair of complex 3D defects using robotic motions generated by the system verifies the effectiveness of the proposed approach and highlights its potential to enhance the practical applicability of AR interfaces in high-precision manufacturing tasks.</div></div>","PeriodicalId":16227,"journal":{"name":"Journal of Manufacturing Systems","volume":"85 ","pages":"Pages 193-206"},"PeriodicalIF":14.2,"publicationDate":"2026-01-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145979109","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

Real-time dynamic integrated process planning and scheduling with reconfigurable manufacturing cells via multi-agent reinforcement learning 基于多智能体强化学习的可重构制造单元实时动态集成工艺规划与调度

IF 14.2 1区工程技术 Q1 ENGINEERING, INDUSTRIAL

Journal of Manufacturing Systems

Pub Date : 2026-01-13 DOI: 10.1016/j.jmsy.2026.01.004

Liang Zheng , Xiaodi Chen , Jianhua Liu , Cunbo Zhuang

Amid the transformation driven by Industry 4.0 and 5.0, manufacturing is rapidly advancing toward greater intelligence and flexibility. Reconfigurable Matrix-structured Manufacturing Systems (RMMS) improve adaptability through dynamic structural and resource reconfiguration, while Integrated Process Planning and Scheduling (IPPS) jointly optimizes process routes and scheduling for optimal resource allocation and responsiveness. This study focuses on Dynamic IPPS with Reconfigurable Manufacturing Cells (DIPPS-RMC) in RMMS, and proposes a real-time scheduling approach based on multi-agent Proximal Policy Optimization (PPO) to reduce average tardiness and enhance system efficiency. A Mixed Integer Linear Programming model is established to address the complexity of process flows and dynamic scheduling, providing a solid theoretical foundation. The scheduling problem is further formulated as a Partially Observable Markov Decision Process to capture the uncertainty and partial observability of real manufacturing environments. To alleviate the credit assignment problem and enhance inter-agent coordination, a delayed reward-sharing mechanism is designed. A multi-agent PPO algorithm with centralized training and decentralized execution is introduced, leveraging parallel environment sampling to improve training efficiency and generalization. Extensive experiments on 270 cases across 27 scenarios show that the proposed method outperforms state-of-the-art multi-agent reinforcement learning algorithms in training speed, generalization, and scheduling performance. Its application to real-world cases further demonstrates effective handling of dynamic job arrivals and RMC breakdowns, validating its robustness and practical utility. These results confirm the method’s effectiveness and applicability in dynamic, complex manufacturing environments, offering an innovative solution for real-time scheduling in RMMS.

在工业4.0和工业5.0的推动下，制造业正迅速向更智能和更灵活的方向发展。可重构矩阵结构制造系统（RMMS）通过动态结构重构和资源重构来提高系统的适应性，而集成工艺规划与调度（IPPS）则通过联合优化工艺路线和调度来实现资源的最优分配和响应。针对RMMS中具有可重构制造单元的动态IPPS (DIPPS-RMC)，提出了一种基于多智能体近端策略优化（PPO）的实时调度方法，以降低平均延迟，提高系统效率。针对复杂的工艺流程和动态调度问题，建立了混合整数线性规划模型，提供了坚实的理论基础。将调度问题进一步表述为部分可观察马尔可夫决策过程，以捕捉真实制造环境的不确定性和部分可观察性。为了缓解信用分配问题，增强代理间的协调能力，设计了一种延迟奖励共享机制。提出了一种集中训练、分散执行的多智能体PPO算法，利用并行环境采样来提高训练效率和泛化能力。在27个场景的270个案例中进行的大量实验表明，所提出的方法在训练速度、泛化和调度性能方面优于最先进的多智能体强化学习算法。它在实际案例中的应用进一步证明了动态工作到达和RMC故障的有效处理，验证了它的鲁棒性和实用性。这些结果证实了该方法在动态、复杂制造环境中的有效性和适用性，为RMMS的实时调度提供了一种创新的解决方案。

{"title":"Real-time dynamic integrated process planning and scheduling with reconfigurable manufacturing cells via multi-agent reinforcement learning","authors":"Liang Zheng , Xiaodi Chen , Jianhua Liu , Cunbo Zhuang","doi":"10.1016/j.jmsy.2026.01.004","DOIUrl":"10.1016/j.jmsy.2026.01.004","url":null,"abstract":"<div><div>Amid the transformation driven by Industry 4.0 and 5.0, manufacturing is rapidly advancing toward greater intelligence and flexibility. Reconfigurable Matrix-structured Manufacturing Systems (RMMS) improve adaptability through dynamic structural and resource reconfiguration, while Integrated Process Planning and Scheduling (IPPS) jointly optimizes process routes and scheduling for optimal resource allocation and responsiveness. This study focuses on Dynamic IPPS with Reconfigurable Manufacturing Cells (DIPPS-RMC) in RMMS, and proposes a real-time scheduling approach based on multi-agent Proximal Policy Optimization (PPO) to reduce average tardiness and enhance system efficiency. A Mixed Integer Linear Programming model is established to address the complexity of process flows and dynamic scheduling, providing a solid theoretical foundation. The scheduling problem is further formulated as a Partially Observable Markov Decision Process to capture the uncertainty and partial observability of real manufacturing environments. To alleviate the credit assignment problem and enhance inter-agent coordination, a delayed reward-sharing mechanism is designed. A multi-agent PPO algorithm with centralized training and decentralized execution is introduced, leveraging parallel environment sampling to improve training efficiency and generalization. Extensive experiments on 270 cases across 27 scenarios show that the proposed method outperforms state-of-the-art multi-agent reinforcement learning algorithms in training speed, generalization, and scheduling performance. Its application to real-world cases further demonstrates effective handling of dynamic job arrivals and RMC breakdowns, validating its robustness and practical utility. These results confirm the method’s effectiveness and applicability in dynamic, complex manufacturing environments, offering an innovative solution for real-time scheduling in RMMS.</div></div>","PeriodicalId":16227,"journal":{"name":"Journal of Manufacturing Systems","volume":"85 ","pages":"Pages 127-154"},"PeriodicalIF":14.2,"publicationDate":"2026-01-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145978606","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