Advanced Engineering Informatics最新文献_第4页

Socio-technical assessment of generative AI integration in architecture, engineering, and construction (AEC) workflows: An empirical study using O*NET occupational taxonomy 建筑、工程和施工（AEC）工作流程中生成式AI集成的社会技术评估：使用O*NET职业分类的实证研究

IF 9.9 1区工程技术 Q1 COMPUTER SCIENCE, ARTIFICIAL INTELLIGENCE

Advanced Engineering Informatics

Pub Date : 2026-04-01 Epub Date: 2026-01-28 DOI: 10.1016/j.aei.2026.104392

Ruoxin Xiong , Yael Netser , Pingbo Tang , Beibei Li , Joonsun Hwang

Generative artificial intelligence (GAI) has the potential to reshape workflows across the Architecture, Engineering, and Construction (AEC) sector. While previous research has offered valuable technical demonstrations and conceptual analyses, empirical evidence quantifying GAI-related impacts across AEC occupations and systematic assessment of adoption readiness remain limited. This study develops a domain-specific socio-technical evaluation framework that provides occupational-level analysis of technical capabilities, social risks, and adoption barriers across thirteen O*NET-defined AEC occupations. Data were collected through a six-month survey of 162 AEC professionals, complemented by six expert interviews and a systematic literature review. The findings reveal: (1) Technical Capability, measured using exposure scores ranging from −1 (low applicability) to +1 (high applicability), shows moderate applicability in design-oriented roles (e.g., architectural drafters: 0.16) and minimal alignment for site-based and manual activities (e.g., construction laborers: −0.89). (2) Social Risks, assessed on a 0–1 scale of concern, identify hallucinations (0.71), data privacy (0.70), and intellectual property issues (0.69) as critical concerns. (3) Socio-Technical Adoption highlights limited technical expertise (26.0%) and uncertain return on investment (16.8%) as primary barriers, while respondents emphasized the need for usage guidelines and standards (29.6%) and targeted training (29.2%) to facilitate responsible integration. Based on these findings, the study outlines strategic priorities for responsible GAI deployment, including AEC-specific standards, targeted workforce training, human-in-the-loop validation mechanisms, and domain-tailored digital infrastructure. The framework and empirical evidence provide a foundation for researchers, practitioners, and policymakers seeking to guide the safe and effective integration of GAI into AEC workflows.

生成式人工智能（GAI）有可能重塑建筑、工程和施工（AEC）领域的工作流程。虽然以前的研究提供了有价值的技术演示和概念分析，但量化人工智能对AEC职业的影响和系统评估采用准备程度的经验证据仍然有限。本研究开发了一个特定领域的社会技术评估框架，提供了13个O* net定义的AEC职业的技术能力、社会风险和采用障碍的职业层面分析。通过对162名AEC专业人员进行为期6个月的调查，并辅以6次专家访谈和系统的文献综述，收集了数据。研究结果显示：(1)技术能力，使用从−1（低适用性）到+1（高适用性）的暴露分数进行测量，显示出在面向设计的角色（例如，建筑起草人：0.16）中的中等适用性，以及对基于现场和手工活动（例如，建筑工人：- 0.89）的最小一致性。(2)社会风险，以0-1的关注程度进行评估，认为幻觉（0.71）、数据隐私（0.70）和知识产权问题（0.69）是关键问题。(3)社会技术采用强调有限的技术专业知识（26.0%）和不确定的投资回报（16.8%）是主要障碍，而受访者强调需要使用指南和标准（29.6%）和有针对性的培训（29.2%）来促进负责任的整合。基于这些发现，该研究概述了负责任的GAI部署的战略重点，包括aec特定标准、有针对性的劳动力培训、人在环验证机制和领域定制的数字基础设施。该框架和经验证据为研究人员、从业者和政策制定者寻求指导GAI安全有效地整合到AEC工作流程中提供了基础。

{"title":"Socio-technical assessment of generative AI integration in architecture, engineering, and construction (AEC) workflows: An empirical study using O*NET occupational taxonomy","authors":"Ruoxin Xiong , Yael Netser , Pingbo Tang , Beibei Li , Joonsun Hwang","doi":"10.1016/j.aei.2026.104392","DOIUrl":"10.1016/j.aei.2026.104392","url":null,"abstract":"<div><div>Generative artificial intelligence (GAI) has the potential to reshape workflows across the Architecture, Engineering, and Construction (AEC) sector. While previous research has offered valuable technical demonstrations and conceptual analyses, empirical evidence quantifying GAI-related impacts across AEC occupations and systematic assessment of adoption readiness remain limited. This study develops a domain-specific socio-technical evaluation framework that provides occupational-level analysis of technical capabilities, social risks, and adoption barriers across thirteen O*NET-defined AEC occupations. Data were collected through a six-month survey of 162 AEC professionals, complemented by six expert interviews and a systematic literature review. The findings reveal: (1) <em>Technical Capability</em>, measured using exposure scores ranging from −1 (low applicability) to +1 (high applicability), shows moderate applicability in design-oriented roles (e.g., architectural drafters: 0.16) and minimal alignment for site-based and manual activities (e.g., construction laborers: −0.89). (2) <em>Social Risks</em>, assessed on a 0–1 scale of concern, identify hallucinations (0.71), data privacy (0.70), and intellectual property issues (0.69) as critical concerns. (3) <em>Socio-Technical Adoption</em> highlights limited technical expertise (26.0%) and uncertain return on investment (16.8%) as primary barriers, while respondents emphasized the need for usage guidelines and standards (29.6%) and targeted training (29.2%) to facilitate responsible integration. Based on these findings, the study outlines strategic priorities for responsible GAI deployment, including AEC-specific standards, targeted workforce training, human-in-the-loop validation mechanisms, and domain-tailored digital infrastructure. The framework and empirical evidence provide a foundation for researchers, practitioners, and policymakers seeking to guide the safe and effective integration of GAI into AEC workflows.</div></div>","PeriodicalId":50941,"journal":{"name":"Advanced Engineering Informatics","volume":"71 ","pages":"Article 104392"},"PeriodicalIF":9.9,"publicationDate":"2026-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146048981","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

MG-LDM: A multimodal guided latent diffusion model with Mamba-based temporal encoding for inverse topological design of tissue engineering skin substitutes MG-LDM：一种基于曼巴时间编码的多模态引导潜在扩散模型，用于组织工程皮肤代用品的逆拓扑设计

IF 9.9 1区工程技术 Q1 COMPUTER SCIENCE, ARTIFICIAL INTELLIGENCE

Advanced Engineering Informatics

Pub Date : 2026-04-01 Epub Date: 2025-12-24 DOI: 10.1016/j.aei.2025.104275

Kaicheng Yu , Wei Zhang , Lihua Lu , Zexue Lin , Ben Chen , Qiang Gao , Guoyin Shang , Peng Zhang

Tissue engineering full-thickness skin substitutes (FSS) with clinically relevant mechanical, structural, and biological characteristics remain a major challenge in regenerative medicine. To address this, we propose a multimodal guided latent diffusion model with Mamba-based temporal encoding (MG-LDM) as a unified inverse design framework for the topology optimization and additive manufacturing of 3D bio-printed FSS. A high-resolution multimodal dataset was constructed, consisting of stress–strain sequences, seven-channel three-dimensional stress field distributions, and extrusion-based 3D printing parameters annotated with cell viability metrics. MG-LDM integrates a U-shaped Mamba encoder for temporal sequence modeling, a densely connected graph convolutional network (DC-GCN) for spatial feature extraction, and a multi-layer perceptron (MLP) encoder for processing manufacturing parameters. These heterogeneous representations are fused via a guided cross-attention mechanism into a unified latent condition, which drives a diffusion-based structure generator. A dual-path topology generation strategy incorporating a neural signed distance function (SDF) ensures geometric continuity and mechanical fidelity. Experimental evaluations indicate that MG-LDM consistently outperforms representative baselines in terms of geometric accuracy, structure–function consistency, and robustness across multiple modalities. Physical validations confirm that MG-LDM enables the generation of biocompatible and mechanically controllable FSS with enhanced structural integrity and regenerative performance, supporting its applicability in personalized regenerative medicine.

组织工程全层皮肤替代品（FSS）具有临床相关的机械、结构和生物学特性，仍然是再生医学的主要挑战。为了解决这个问题，我们提出了一种基于曼巴时间编码（MG-LDM）的多模态引导潜在扩散模型，作为生物3D打印FSS拓扑优化和增材制造的统一逆设计框架。构建了一个高分辨率的多模态数据集，包括应力-应变序列、七通道三维应力场分布和基于挤压的3D打印参数，并标注了细胞活力指标。MG-LDM集成了用于时间序列建模的u形曼巴编码器，用于空间特征提取的密集连接图卷积网络（DC-GCN）和用于处理制造参数的多层感知器（MLP）编码器。这些异质表征通过一个引导的交叉注意机制融合成一个统一的潜在条件，驱动一个基于扩散的结构生成器。基于神经符号距离函数（SDF）的双路径拓扑生成策略保证了几何连续性和机械保真度。实验评估表明，MG-LDM在几何精度、结构-功能一致性和多模态鲁棒性方面始终优于代表性基线。物理验证证实MG-LDM能够生成生物相容性和机械可控的FSS，具有增强的结构完整性和再生性能，支持其在个性化再生医学中的适用性。

{"title":"MG-LDM: A multimodal guided latent diffusion model with Mamba-based temporal encoding for inverse topological design of tissue engineering skin substitutes","authors":"Kaicheng Yu , Wei Zhang , Lihua Lu , Zexue Lin , Ben Chen , Qiang Gao , Guoyin Shang , Peng Zhang","doi":"10.1016/j.aei.2025.104275","DOIUrl":"10.1016/j.aei.2025.104275","url":null,"abstract":"<div><div>Tissue engineering full-thickness skin substitutes (FSS) with clinically relevant mechanical, structural, and biological characteristics remain a major challenge in regenerative medicine. To address this, we propose a multimodal guided latent diffusion model with Mamba-based temporal encoding (MG-LDM) as a unified inverse design framework for the topology optimization and additive manufacturing of 3D bio-printed FSS. A high-resolution multimodal dataset was constructed, consisting of stress–strain sequences, seven-channel three-dimensional stress field distributions, and extrusion-based 3D printing parameters annotated with cell viability metrics. MG-LDM integrates a U-shaped Mamba encoder for temporal sequence modeling, a densely connected graph convolutional network (DC-GCN) for spatial feature extraction, and a multi-layer perceptron (MLP) encoder for processing manufacturing parameters. These heterogeneous representations are fused via a guided cross-attention mechanism into a unified latent condition, which drives a diffusion-based structure generator. A dual-path topology generation strategy incorporating a neural signed distance function (SDF) ensures geometric continuity and mechanical fidelity. Experimental evaluations indicate that MG-LDM consistently outperforms representative baselines in terms of geometric accuracy, structure–function consistency, and robustness across multiple modalities. Physical validations confirm that MG-LDM enables the generation of biocompatible and mechanically controllable FSS with enhanced structural integrity and regenerative performance, supporting its applicability in personalized regenerative medicine.</div></div>","PeriodicalId":50941,"journal":{"name":"Advanced Engineering Informatics","volume":"71 ","pages":"Article 104275"},"PeriodicalIF":9.9,"publicationDate":"2026-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145841894","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 damage-integrated image-point cloud registration network for characterizing macro-meso hysteretic features of semi-rigid precast concrete beam-column joints under strong earthquakes 强震作用下半刚性预制混凝土梁柱节点宏细观滞回特征的损伤综合像点云配准网络

IF 9.9 1区工程技术 Q1 COMPUTER SCIENCE, ARTIFICIAL INTELLIGENCE

Advanced Engineering Informatics

Pub Date : 2026-04-01 Epub Date: 2025-12-26 DOI: 10.1016/j.aei.2025.104260

Dianyou Yu , Zheng He , Yuexuan Yu

Semi-rigid precast concrete (PC) connections can offer good design adaptability to provide desired strength, stiffness, and energy dissipation. When subjected to earthquake-excited cyclic loading, they display prominent non-integral global deformation, complex local multi-scale surface features, and temporal multi-valued attributes. To overcome the challenge of establishing the mapping relation between surface features and internal damage states in semi-rigid PC joints, we propose HysD-I2PC, a novel registration network. This system integrates three key components: an image-point cloud registration module, a cross-dimensional fusion mechanism, and a damage prediction module, enabling end-to-end pose estimation and temporal damage prediction. A comprehensive dataset consisting of 2016 samples from four tested semi-rigid PC joints and 92 simulated ones, as well as the public KITTI, is used for the comparative analysis with DeepI2P, CorrI2P, VP2P, and CoFil2P. The outcome of this study highlights the rationality of the architecture of HysD-I2PC and its superior performance in describing the prominent macro-meso hysteresis features and predicting temporal damage of the joints under reversed cyclic loading. HysD-I2PC possesses ResNet-Transformer and Point Transformer in the image and point cloud branches, respectively, a cross-dimensional feature fusion mechanism and a damage-driven temporal prediction branch. It is proved to be optimal from the ablation experiments with a loss function reflecting comprehensively the contributions from feature similarity, differential pose, and hysteretic damage. For the generated dataset, the errors of relative rotation, relative translation, and damage prediction of the proposed registration network are only 1.07 ± 0.39°, 1.13 ± 0.72 mm, and 6.81 ± 5.40 %, respectively.

半刚性预制混凝土（PC）连接可以提供良好的设计适应性，以提供所需的强度、刚度和耗能。在地震激发循环荷载作用下，它们表现出突出的非整体变形、复杂的局部多尺度表面特征和时间多值属性。为了克服半刚性PC节点表面特征与内部损伤状态之间映射关系的挑战，我们提出了一种新的HysD-I2PC配准网络。该系统集成了三个关键组件：图像点云配准模块、跨维融合机制和损伤预测模块，实现端到端姿态估计和时间损伤预测。利用2016年4个半刚性PC接头和92个模拟接头以及公开的KITTI样本的综合数据集，与DeepI2P、CorrI2P、VP2P和CoFil2P进行比较分析。本研究结果突出了HysD-I2PC结构的合理性，以及它在描述显著的宏细观迟滞特征和预测反循环荷载作用下关节时间损伤方面的优越性能。HysD-I2PC在图像和点云分支中分别具有ResNet-Transformer和Point Transformer、跨维特征融合机制和损伤驱动时间预测分支。通过烧蚀实验证明该方法是最优的，其损失函数综合反映了特征相似度、差分位姿和迟滞损伤的贡献。对于生成的数据集，所提出的配准网络的相对旋转、相对平移和损伤预测误差分别仅为1.07±0.39°、1.13±0.72 mm和6.81±5.40%。

{"title":"A damage-integrated image-point cloud registration network for characterizing macro-meso hysteretic features of semi-rigid precast concrete beam-column joints under strong earthquakes","authors":"Dianyou Yu , Zheng He , Yuexuan Yu","doi":"10.1016/j.aei.2025.104260","DOIUrl":"10.1016/j.aei.2025.104260","url":null,"abstract":"<div><div>Semi-rigid precast concrete (PC) connections can offer good design adaptability to provide desired strength, stiffness, and energy dissipation. When subjected to earthquake-excited cyclic loading, they display prominent non-integral global deformation, complex local multi-scale surface features, and temporal multi-valued attributes. To overcome the challenge of establishing the mapping relation between surface features and internal damage states in semi-rigid PC joints, we propose HysD-I2PC, a novel registration network. This system integrates three key components: an image-point cloud registration module, a cross-dimensional fusion mechanism, and a damage prediction module, enabling end-to-end pose estimation and temporal damage prediction. A comprehensive dataset consisting of 2016 samples from four tested semi-rigid PC joints and 92 simulated ones, as well as the public KITTI, is used for the comparative analysis with DeepI2P, CorrI2P, VP2P, and CoFil2P. The outcome of this study highlights the rationality of the architecture of HysD-I2PC and its superior performance in describing the prominent macro-meso hysteresis features and predicting temporal damage of the joints under reversed cyclic loading. HysD-I2PC possesses ResNet-Transformer and Point Transformer in the image and point cloud branches, respectively, a cross-dimensional feature fusion mechanism and a damage-driven temporal prediction branch. It is proved to be optimal from the ablation experiments with a loss function reflecting comprehensively the contributions from feature similarity, differential pose, and hysteretic damage. For the generated dataset, the errors of relative rotation, relative translation, and damage prediction of the proposed registration network are only 1.07 ± 0.39°, 1.13 ± 0.72 mm, and 6.81 ± 5.40 %, respectively.</div></div>","PeriodicalId":50941,"journal":{"name":"Advanced Engineering Informatics","volume":"71 ","pages":"Article 104260"},"PeriodicalIF":9.9,"publicationDate":"2026-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145841966","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

Integrating context awareness and knowledge graphs for enhanced knowledge recommendation in manufacturing process planning 集成上下文感知和知识图谱，增强制造工艺规划中的知识推荐

IF 9.9 1区工程技术 Q1 COMPUTER SCIENCE, ARTIFICIAL INTELLIGENCE

Advanced Engineering Informatics

Pub Date : 2026-04-01 Epub Date: 2026-01-12 DOI: 10.1016/j.aei.2026.104324

Zhenyong Wu , Hanqing Wu , Jianxin Tan , Ying Qian , Lina He , Mark Goh

Complex process planning is a knowledge-intensive task requiring effective knowledge reuse among distributed teams and engineers. As efficient knowledge acquisition and rapid retrieval of relevant knowledge persist as key challenges in manufacturing process planning, this paper proposes a knowledge graph-based representation model aimed at facilitating more effective knowledge utilization throughout the process planning lifecycle. This study develops a generative adversarial network-based context-aware recommendation system to support effective knowledge acquisition and reuse by delivering recommendations customized to engineers’ specific process requirements. The proposed framework comprehensively captures and integrates contextual information, streamlining the knowledge retrieval process. The practical applicability and performance of the proposed method is validated through a case study, achieving an F1-score of 0.519 and reducing knowledge retrieval time by more than 50%.

复杂过程规划是一项知识密集型任务，需要在分布式团队和工程师之间进行有效的知识重用。针对制造工艺规划中知识的高效获取和快速检索一直是关键挑战，本文提出了一种基于知识图的表示模型，旨在促进整个工艺规划生命周期中知识的更有效利用。本研究开发了一个基于生成对抗网络的上下文感知推荐系统，通过提供针对工程师特定流程需求的定制推荐，支持有效的知识获取和重用。该框架全面地捕获和集成了上下文信息，简化了知识检索过程。通过实例验证了该方法的实用性和性能，f1得分为0.519，知识检索时间减少了50%以上。

引用次数: 0

A 3D geometry adaptive approach for fusing heterogeneous point clouds in construction sites 建筑工地异构点云的三维几何自适应融合方法

IF 9.9 1区工程技术 Q1 COMPUTER SCIENCE, ARTIFICIAL INTELLIGENCE

Advanced Engineering Informatics

Pub Date : 2026-04-01 Epub Date: 2026-02-05 DOI: 10.1016/j.aei.2026.104398

Yujie Lu , Tao Zhong , Haoyu Deng , Shuo Wang , Chuan Yang , Xianzhong Zhao

High-fidelity 3D modeling of large-scale construction sites requires the fusion of multi-source point clouds to satisfy the diverse requirements of various engineering tasks. A primary challenge in this process is the non-uniform quality of the data, as the relative precision between different sources varies spatially across the site. This variability can cause holistic fusion approaches to erroneously discard high-quality local data. To address this, the presented framework reframes the global fusion challenge into a more tractable problem of local partitioning and optimization. A novel hierarchical partitioning method creates local units that are geometrically adaptive to building components, enhancing both geometric continuity and processing efficiency. The methodology consists of three sequential stages: (1) high-precision registration of all source data; (2) hierarchical partitioning at both architectural and sub-component levels to create basic fusion units; and (3) multi-strategy local filtering based on data quality within each unit. The framework was validated on a challenging case involving the fusion of multi-source, image-based point clouds from a super-tall building site. The results demonstrate a precision improvement of 6.8% for the entire site and up to 31.9% for detailed regions compared to the simply merged data. The methodology enhances the integrity of as-built models and improves the representation of surface textures and structural dimensions, providing a reliable data foundation for downstream construction management tasks such as spatial progress monitoring.

大型建筑工地的高保真三维建模需要多源点云的融合，以满足各种工程任务的多样化需求。这个过程中的一个主要挑战是数据的质量不一致，因为不同来源之间的相对精度在整个站点的空间上是不同的。这种可变性可能导致整体融合方法错误地丢弃高质量的本地数据。为了解决这一问题，本文提出的框架将全局融合问题重构为一个更易于处理的局部划分和优化问题。一种新颖的分层划分方法创建了几何上适应建筑构件的局部单元，提高了几何连续性和处理效率。该方法包括三个连续阶段：(1)所有源数据的高精度配准；(2)在体系结构和子组件层面进行分层划分，创建基本融合单元；(3)基于各单元内数据质量的多策略局部滤波。该框架在一个具有挑战性的案例中得到了验证，该案例涉及来自超高建筑工地的多源、基于图像的点云融合。结果表明，与简单合并的数据相比，整个站点的精度提高了6.8%，详细区域的精度提高了31.9%。该方法提高了竣工模型的完整性，改善了表面纹理和结构维度的表达，为下游施工管理任务（如空间进度监测）提供了可靠的数据基础。

{"title":"A 3D geometry adaptive approach for fusing heterogeneous point clouds in construction sites","authors":"Yujie Lu , Tao Zhong , Haoyu Deng , Shuo Wang , Chuan Yang , Xianzhong Zhao","doi":"10.1016/j.aei.2026.104398","DOIUrl":"10.1016/j.aei.2026.104398","url":null,"abstract":"<div><div>High-fidelity 3D modeling of large-scale construction sites requires the fusion of multi-source point clouds to satisfy the diverse requirements of various engineering tasks. A primary challenge in this process is the non-uniform quality of the data, as the relative precision between different sources varies spatially across the site. This variability can cause holistic fusion approaches to erroneously discard high-quality local data. To address this, the presented framework reframes the global fusion challenge into a more tractable problem of local partitioning and optimization. A novel hierarchical partitioning method creates local units that are geometrically adaptive to building components, enhancing both geometric continuity and processing efficiency. The methodology consists of three sequential stages: (1) high-precision registration of all source data; (2) hierarchical partitioning at both architectural and sub-component levels to create basic fusion units; and (3) multi-strategy local filtering based on data quality within each unit. The framework was validated on a challenging case involving the fusion of multi-source, image-based point clouds from a super-tall building site. The results demonstrate a precision improvement of 6.8% for the entire site and up to 31.9% for detailed regions compared to the simply merged data. The methodology enhances the integrity of as-built models and improves the representation of surface textures and structural dimensions, providing a reliable data foundation for downstream construction management tasks such as spatial progress monitoring.</div></div>","PeriodicalId":50941,"journal":{"name":"Advanced Engineering Informatics","volume":"71 ","pages":"Article 104398"},"PeriodicalIF":9.9,"publicationDate":"2026-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146187868","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

ExpertPlanner: A mixture-of-experts transformer language model for decomposing construction look-ahead plan tasks from long-term master schedules ExpertPlanner：一个混合专家转换语言模型，用于从长期主计划中分解施工预判计划任务

IF 9.9 1区工程技术 Q1 COMPUTER SCIENCE, ARTIFICIAL INTELLIGENCE

Advanced Engineering Informatics

Pub Date : 2026-04-01 Epub Date: 2026-02-10 DOI: 10.1016/j.aei.2026.104439

Yoonhwa Jung , Mani Golparvar-Fard

A well-managed look-ahead plan, consistently reconciled with the master schedule, ensures that everyone on a construction project has the right information to do their part efficiently. However, manually creating look-ahead plans that maintain alignment with long-term milestones creates an administrative overload that often results in poorly managed task-level planning. To address this, this paper presents ExpertPlanner, a domain-specific Mixture-of-Experts (MoE) language model integrated with Curriculum Learning that automatically decomposes high-level master schedule activities into granular look-ahead planning tasks. Unlike traditional models, our MoE architecture features a UniversalExpert that processes global context in parallel with specialized experts, enhancing load balancing and generalization. Using real-world project data, the performance of ExpertPlanner outperforms baselines and generic large language model (LLM), achieving a BLEU@4 of 71.2 and an average score of 82.4. Our results demonstrate the potential of the proposed method to assist superintendents in generating look-ahead planning tasks. By automating this transformation, our model acts as a vital bridge between high-level project strategy and daily operational execution. This direct connection promotes consistency and visibility, ensures that long-term and short-term plans remain reconciled and traceable, and enhances communication. Furthermore, the model maintains consistency with master schedule activities while allowing users to make manual adjustments easily, while remaining lightweight and cost-efficient.

一个管理良好的前瞻性计划，始终与总进度保持一致，确保建筑项目中的每个人都有正确的信息来有效地完成他们的工作。然而，手动创建与长期里程碑保持一致的预见性计划会造成管理负担过重，通常会导致管理不善的任务级计划。为了解决这个问题，本文提出了ExpertPlanner，这是一个领域特定的专家混合（MoE）语言模型，它与课程学习集成在一起，自动将高级主计划活动分解为细粒度的前瞻性计划任务。与传统模型不同，我们的MoE架构具有一个UniversalExpert，它与专业专家并行处理全局上下文，增强了负载平衡和泛化。使用实际项目数据，ExpertPlanner的性能优于基线和通用大型语言模型（LLM），得分BLEU@4为71.2，平均得分为82.4。我们的结果证明了所提出的方法在帮助管理者生成前瞻性规划任务方面的潜力。通过自动化此转换，我们的模型充当了高层项目策略和日常操作执行之间的重要桥梁。这种直接联系促进了一致性和可见性，确保长期和短期计划保持协调和可追溯性，并增强了沟通。此外，该模型与主计划活动保持一致，同时允许用户轻松地进行手动调整，同时保持轻量级和成本效益。

{"title":"ExpertPlanner: A mixture-of-experts transformer language model for decomposing construction look-ahead plan tasks from long-term master schedules","authors":"Yoonhwa Jung , Mani Golparvar-Fard","doi":"10.1016/j.aei.2026.104439","DOIUrl":"10.1016/j.aei.2026.104439","url":null,"abstract":"<div><div>A well-managed look-ahead plan, consistently reconciled with the master schedule, ensures that everyone on a construction project has the right information to do their part efficiently. However, manually creating look-ahead plans that maintain alignment with long-term milestones creates an administrative overload that often results in poorly managed task-level planning. To address this, this paper presents <span>ExpertPlanner</span>, a domain-specific Mixture-of-Experts (MoE) language model integrated with Curriculum Learning that automatically decomposes high-level master schedule activities into granular look-ahead planning tasks. Unlike traditional models, our MoE architecture features a UniversalExpert that processes global context in parallel with specialized experts, enhancing load balancing and generalization. Using real-world project data, the performance of <span>ExpertPlanner</span> outperforms baselines and generic large language model (LLM), achieving a BLEU@4 of 71.2 and an average score of 82.4. Our results demonstrate the potential of the proposed method to assist superintendents in generating look-ahead planning tasks. By automating this transformation, our model acts as a vital bridge between high-level project strategy and daily operational execution. This direct connection promotes consistency and visibility, ensures that long-term and short-term plans remain reconciled and traceable, and enhances communication. Furthermore, the model maintains consistency with master schedule activities while allowing users to make manual adjustments easily, while remaining lightweight and cost-efficient.</div></div>","PeriodicalId":50941,"journal":{"name":"Advanced Engineering Informatics","volume":"71 ","pages":"Article 104439"},"PeriodicalIF":9.9,"publicationDate":"2026-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146188694","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

Explainable wavelet-scalogram learning for quasi-stationary faults in automotive DC motors using AERIS-Wave 基于AERIS-Wave的汽车直流电机准平稳故障的可解释小波尺度学习

IF 9.9 1区工程技术 Q1 COMPUTER SCIENCE, ARTIFICIAL INTELLIGENCE

Advanced Engineering Informatics

Pub Date : 2026-04-01 Epub Date: 2026-02-03 DOI: 10.1016/j.aei.2026.104394

Willy Dharmawan , Nur Hamid , Haitham Saleh , Amalia Irma Nurwidya , Peni Laksmita Widati

Noise diagnostics in automotive brushed DC motors are crucial for ensuring reliability and compliance with strict manufacturing standards. Traditional human-based inspection methods are often subjective and inconsistent, necessitating automated, data-driven solutions. This study introduces an explainable wavelet-based classification framework for diagnosing quasi-stationary motor faults using accelerometer data. The framework employs the Continuous Wavelet Transform (CWT) to extract rich time–frequency features, including scalograms, ridge trajectories, periodograms, and statistical descriptors, serving as discriminative representations of mechanical behavior. These features are evaluated using both classical machine learning algorithms (Random Forest, XGBoost) and deep learning architectures such as ResNet18, VGG16, CNN-LSTM, CNN-GRU, and WaveNet. Building upon these baselines, the proposed AERIS-Wave (Attention-Enhanced Residual Interpretable Scalogram Network) integrates multi-level attention, LayerScale normalization, and explainable-AI components (Integrated Gradients, Grad-CAM, SHAP) to visualize spectral contributions that drive decisions. Experimental results show that AERIS-Wave achieves 97.72% accuracy and an AUC of 0.9991, surpassing all benchmark models, including WaveNet and ResNet18. The findings confirm that wavelet-based representations, combined with interpretable deep learning, enable high-precision, explainable, and scalable fault classification suitable for real-time quality control in industrial environments.

汽车有刷直流电机的噪声诊断对于确保可靠性和符合严格的制造标准至关重要。传统的人工检测方法往往是主观的和不一致的，需要自动化的、数据驱动的解决方案。本文介绍了一种基于小波的可解释分类框架，用于利用加速度计数据诊断准静止电机故障。该框架采用连续小波变换（CWT）提取丰富的时频特征，包括尺度图、脊轨迹、周期图和统计描述符，作为机械行为的判别表示。这些特征使用经典机器学习算法（Random Forest, XGBoost）和深度学习架构（如ResNet18， VGG16, CNN-LSTM， CNN-GRU和WaveNet）进行评估。在这些基线的基础上，提出的AERIS-Wave（注意力增强残余可解释尺度网络）集成了多层次注意力、LayerScale归一化和可解释的人工智能组件（集成梯度、gradcam、SHAP），以可视化驱动决策的光谱贡献。实验结果表明，AERIS-Wave的准确率为97.72%，AUC为0.9991，优于WaveNet和ResNet18等所有基准模型。研究结果证实，基于小波的表示与可解释的深度学习相结合，可以实现高精度、可解释和可扩展的故障分类，适用于工业环境中的实时质量控制。

{"title":"Explainable wavelet-scalogram learning for quasi-stationary faults in automotive DC motors using AERIS-Wave","authors":"Willy Dharmawan , Nur Hamid , Haitham Saleh , Amalia Irma Nurwidya , Peni Laksmita Widati","doi":"10.1016/j.aei.2026.104394","DOIUrl":"10.1016/j.aei.2026.104394","url":null,"abstract":"<div><div>Noise diagnostics in automotive brushed DC motors are crucial for ensuring reliability and compliance with strict manufacturing standards. Traditional human-based inspection methods are often subjective and inconsistent, necessitating automated, data-driven solutions. This study introduces an explainable wavelet-based classification framework for diagnosing quasi-stationary motor faults using accelerometer data. The framework employs the Continuous Wavelet Transform (CWT) to extract rich time–frequency features, including scalograms, ridge trajectories, periodograms, and statistical descriptors, serving as discriminative representations of mechanical behavior. These features are evaluated using both classical machine learning algorithms (Random Forest, XGBoost) and deep learning architectures such as ResNet18, VGG16, CNN-LSTM, CNN-GRU, and WaveNet. Building upon these baselines, the proposed AERIS-Wave (Attention-Enhanced Residual Interpretable Scalogram Network) integrates multi-level attention, LayerScale normalization, and explainable-AI components (Integrated Gradients, Grad-CAM, SHAP) to visualize spectral contributions that drive decisions. Experimental results show that AERIS-Wave achieves 97.72% accuracy and an AUC of 0.9991, surpassing all benchmark models, including WaveNet and ResNet18. The findings confirm that wavelet-based representations, combined with interpretable deep learning, enable high-precision, explainable, and scalable fault classification suitable for real-time quality control in industrial environments.</div></div>","PeriodicalId":50941,"journal":{"name":"Advanced Engineering Informatics","volume":"71 ","pages":"Article 104394"},"PeriodicalIF":9.9,"publicationDate":"2026-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146188699","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 knowledge-formalization and vision-language-model system for fire emergency decision support 面向火灾应急决策支持的知识形式化视觉语言模型系统

IF 9.9 1区工程技术 Q1 COMPUTER SCIENCE, ARTIFICIAL INTELLIGENCE

Advanced Engineering Informatics

Pub Date : 2026-04-01 Epub Date: 2026-02-10 DOI: 10.1016/j.aei.2026.104409

Jianyuan Tao , Ping Wang , Hanfeng Jiang , Yu Han

Fire incident reports and on-site images contain rich but unstructured fire-safety engineering knowledge that requires explicit formalization to support knowledge-intensive decision-making. The large volume and multimodal nature of such evidence make it difficult to transform it into computable and reusable knowledge representations. This paper proposes a hybrid decision support system, KG-DS, that formalizes fire-safety engineering knowledge through a Fire Incident Knowledge Graph (FIKG) and integrates it with VLM-based multimodal reasoning for fire emergency decision support. The system comprises two modules: (1) a Relation-Oriented Joint Triples Extraction Model (RO-JTEM), which serves as a knowledge-formalization mechanism by extracting

<

subject-relation-object

>

event triples from fire investigation texts to construct a structured FIKG; and (2) a VLM-based reasoning module that combines graph retrieval with multimodal perception to form interpretable Perception–Thought–Action loops for autonomous strategy generation. RO-JTEM achieves an F1-score of 85.2% in triple extraction, while KG-DS attains an average similarity of 81.4% with expert-generated strategies. These results demonstrate that coupling explicit engineering knowledge representations with multimodal computational reasoning provides a reliable and explainable mechanism for supporting knowledge-intensive fire emergency decision-making.

火灾事故报告和现场图像包含丰富但非结构化的消防安全工程知识，需要明确形式化以支持知识密集型决策。此类证据的大容量和多模态性质使得将其转换为可计算和可重用的知识表示变得困难。本文提出了一种混合决策支持系统KG-DS，该系统通过火灾事件知识图（FIKG）形式化消防安全工程知识，并将其与基于vmm的多模态推理相结合，用于火灾应急决策支持。该系统包括两个模块：(1)面向关系的联合三元组抽取模型（RO-JTEM），作为知识形式化机制，从火灾调查文本中抽取“主体-关系-客体”事件三元组，构建结构化FIKG；(2)基于vlm的推理模块，该模块将图检索与多模态感知相结合，形成可解释的感知-思考-行动循环，用于自主策略生成。RO-JTEM在三次提取中获得了85.2%的f1分数，而KG-DS与专家生成策略的平均相似度为81.4%。这些结果表明，将显式工程知识表示与多模态计算推理相结合，为支持知识密集型火灾应急决策提供了可靠且可解释的机制。

{"title":"A knowledge-formalization and vision-language-model system for fire emergency decision support","authors":"Jianyuan Tao , Ping Wang , Hanfeng Jiang , Yu Han","doi":"10.1016/j.aei.2026.104409","DOIUrl":"10.1016/j.aei.2026.104409","url":null,"abstract":"<div><div>Fire incident reports and on-site images contain rich but unstructured fire-safety engineering knowledge that requires explicit formalization to support knowledge-intensive decision-making. The large volume and multimodal nature of such evidence make it difficult to transform it into computable and reusable knowledge representations. This paper proposes a hybrid decision support system, KG-DS, that formalizes fire-safety engineering knowledge through a Fire Incident Knowledge Graph (FIKG) and integrates it with VLM-based multimodal reasoning for fire emergency decision support. The system comprises two modules: (1) a Relation-Oriented Joint Triples Extraction Model (RO-JTEM), which serves as a knowledge-formalization mechanism by extracting <span><math><mo><</mo></math></span>subject-relation-object<span><math><mo>></mo></math></span> event triples from fire investigation texts to construct a structured FIKG; and (2) a VLM-based reasoning module that combines graph retrieval with multimodal perception to form interpretable Perception–Thought–Action loops for autonomous strategy generation. RO-JTEM achieves an F1-score of 85.2% in triple extraction, while KG-DS attains an average similarity of 81.4% with expert-generated strategies. These results demonstrate that coupling explicit engineering knowledge representations with multimodal computational reasoning provides a reliable and explainable mechanism for supporting knowledge-intensive fire emergency decision-making.</div></div>","PeriodicalId":50941,"journal":{"name":"Advanced Engineering Informatics","volume":"71 ","pages":"Article 104409"},"PeriodicalIF":9.9,"publicationDate":"2026-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146188817","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

FGAT: A states labeling method with Fuzzy Graph Attention Network for industrial protocol reverse engineering 基于模糊图注意网络的工业协议逆向工程状态标注方法

IF 9.9 1区工程技术 Q1 COMPUTER SCIENCE, ARTIFICIAL INTELLIGENCE

Advanced Engineering Informatics

Pub Date : 2026-04-01 Epub Date: 2026-02-10 DOI: 10.1016/j.aei.2026.104425

Weikang Zhou , Guangfu Ma , Nan Zhou , Xiaojun Liang , Cunnian Gao , Wenfeng Deng , Chunhua Yang , Weihua Gui

The deep integration between Information Technology (IT) networks and Operational Technology (OT) networks introduced hybrid traffic patterns and increased link complexity, thereby exacerbating challenges in protocol configuration, conversion, and security assurance. As the foundational prerequisite for these processes, protocol analysis urgently requires transformation from current manual operations to automated, intelligent paradigms capable of performing efficient and accurate reverse analysis on increasingly complex Industrial Internet traffic. Conventional reverse engineering methods demonstrate limited effectiveness in message state labeling for industrial protocols due to their proprietary nature and sequence transparency characteristics, while also failing to adequately capture implicit inter-message relationships. To address these challenges, we propose FGAT, a fuzzy logic aggregation graph neural network based on attention mechanism for industrial protocol message state labeling. Our approach transforms protocol message data into graph structures and performs feature aggregation through attention mechanisms combined with fuzzy logic, thereby incorporating strong relational inductive biases between messages. We conducted experiments on FGAT using industrial traffic data of seven different protocols, and demonstrated the superiority of our method by comparing the classification accuracy of FGAT (97.16%) with the traditional graph neural aggregation method (85.61%/86.96%/87.08%), the latest fuzzy graph neural network methods (81.70%/79.53%), traditional unsupervised deep learning methods (89.98%/69.57%/95.16%/85.98%/86.89%), and the message state labeling method of traditional protocol reverse technology (47.50%/54.26%/71.49%).

信息技术（IT）网络与运营技术（OT）网络的深度融合引入了混合流量模式，增加了链路复杂性，从而加剧了协议配置、转换和安全保障方面的挑战。协议分析作为这些过程的基础前提，迫切需要从目前的人工操作转变为自动化、智能的范式，能够对日益复杂的工业互联网流量进行高效、准确的反向分析。传统的逆向工程方法由于其专有性质和序列透明特性，在工业协议的消息状态标记方面的有效性有限，同时也无法充分捕获隐含的消息间关系。为了解决这些问题，我们提出了一种基于关注机制的模糊逻辑聚合图神经网络FGAT，用于工业协议消息状态标注。我们的方法将协议消息数据转换为图结构，并通过注意机制结合模糊逻辑进行特征聚合，从而结合消息之间的强关系归纳偏差。利用7种不同协议的工业交通数据对FGAT进行了实验，结果表明，FGAT的分类准确率（97.16%）与传统的图神经聚合方法（85.61%/86.96%/87.08%）、最新的模糊图神经网络方法（81.70%/79.53%）、传统的无监督深度学习方法（89.98%/69.57%/95.16%/85.98%/86.89%）、而传统协议反向技术的消息状态标注方法（47.50%/54.26%/71.49%）。

{"title":"FGAT: A states labeling method with Fuzzy Graph Attention Network for industrial protocol reverse engineering","authors":"Weikang Zhou , Guangfu Ma , Nan Zhou , Xiaojun Liang , Cunnian Gao , Wenfeng Deng , Chunhua Yang , Weihua Gui","doi":"10.1016/j.aei.2026.104425","DOIUrl":"10.1016/j.aei.2026.104425","url":null,"abstract":"<div><div>The deep integration between Information Technology (IT) networks and Operational Technology (OT) networks introduced hybrid traffic patterns and increased link complexity, thereby exacerbating challenges in protocol configuration, conversion, and security assurance. As the foundational prerequisite for these processes, protocol analysis urgently requires transformation from current manual operations to automated, intelligent paradigms capable of performing efficient and accurate reverse analysis on increasingly complex Industrial Internet traffic. Conventional reverse engineering methods demonstrate limited effectiveness in message state labeling for industrial protocols due to their proprietary nature and sequence transparency characteristics, while also failing to adequately capture implicit inter-message relationships. To address these challenges, we propose FGAT, a fuzzy logic aggregation graph neural network based on attention mechanism for industrial protocol message state labeling. Our approach transforms protocol message data into graph structures and performs feature aggregation through attention mechanisms combined with fuzzy logic, thereby incorporating strong relational inductive biases between messages. We conducted experiments on FGAT using industrial traffic data of seven different protocols, and demonstrated the superiority of our method by comparing the classification accuracy of FGAT (97.16%) with the traditional graph neural aggregation method (85.61%/86.96%/87.08%), the latest fuzzy graph neural network methods (81.70%/79.53%), traditional unsupervised deep learning methods (89.98%/69.57%/95.16%/85.98%/86.89%), and the message state labeling method of traditional protocol reverse technology (47.50%/54.26%/71.49%).</div></div>","PeriodicalId":50941,"journal":{"name":"Advanced Engineering Informatics","volume":"71 ","pages":"Article 104425"},"PeriodicalIF":9.9,"publicationDate":"2026-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146188819","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

From product specifications to smart manufacturing: Generative AI-driven process innovation for BOM digitization and its applications 从产品规格到智能制造：生成式人工智能驱动的BOM数字化流程创新及其应用

IF 9.9 1区工程技术 Q1 COMPUTER SCIENCE, ARTIFICIAL INTELLIGENCE

Advanced Engineering Informatics

Pub Date : 2026-04-01 Epub Date: 2026-02-11 DOI: 10.1016/j.aei.2026.104451

Yu-Chi Lin, Jr-Fong Dang

As global manufacturing advances toward digital transformation and intelligent operations, seamless integration between product specification documents and Bills of Materials (BOM) has become increasingly critical. Conventional BOM creation processes rely heavily on manual interpretation. This often leads to inconsistencies in part numbering, redundant records, and extended engineering cycles, especially when documentation is unstructured and multilingual. To address these challenges, this study proposes a generative AI–driven framework that integrates large language models (LLMs) with retrieval-augmented generation (RAG) for intelligent BOM digitization. The framework establishes a closed-loop pipeline. It converts unstructured product datasheets into rule-validated and semantically consistent part numbers through multi-stage document parsing, knowledge retrieval, and constrained decoding. A human-in-the-loop verification mechanism is incorporated to ensure semantic precision and traceability. By leveraging the multimodal capabilities of the Qwen model, the proposed framework effectively parses and understands bilingual (Chinese–English) technical documents. This enhances cross-lingual semantic comprehension and data accuracy. The validated outputs are consolidated into an executable Electrical and Digital BOM (x-EDBOM). This x-EDBOM serves as a unified source of truth across Product Lifecycle Management (PLM), Enterprise Resource Planning (ERP), and Manufacturing Execution Systems (MES). Empirical validation at a mid-sized electronics manufacturer demonstrates a 12.5× reduction in processing time (25 to 2 min), an average accuracy of 91%. It also indicates estimated monthly cost savings of approximately USD 0.7 million. Furthermore, the x-EDBOM enables comparative analysis of identical components across multiple suppliers. It integrates their price, lead-time, and specification data within a unified system, thereby supporting data-driven sourcing optimization and supplier management. Beyond efficiency improvements, the proposed framework supports cross-lingual data normalization, supplier benchmarking, and post-merger system harmonization. It provides a scalable foundation for intelligent product data management and smart manufacturing transformation.

随着全球制造业向数字化转型和智能运营迈进，产品规格文件和物料清单（BOM）之间的无缝集成变得越来越重要。传统的BOM创建过程严重依赖于人工解释。这通常会导致零件编号不一致、冗余记录和延长的工程周期，特别是当文档是非结构化和多语言的时候。为了应对这些挑战，本研究提出了一个生成式人工智能驱动框架，该框架将大型语言模型（llm）与检索增强生成（RAG）集成在一起，用于智能BOM数字化。该框架建立了一个闭环管道。它通过多阶段文档解析、知识检索和约束解码将非结构化产品数据表转换为规则验证和语义一致的部件号。引入了人在环验证机制，以确保语义精确性和可追溯性。通过利用Qwen模型的多模式功能，所提出的框架可以有效地解析和理解双语（中英）技术文档。这增强了跨语言语义理解和数据准确性。经过验证的输出被合并到可执行的电气和数字BOM （x-EDBOM）中。该x-EDBOM作为跨产品生命周期管理（PLM）、企业资源规划（ERP）和制造执行系统（MES）的统一数据源。在一家中型电子制造商的经验验证表明，处理时间（25至2分钟）减少了12.5倍，平均精度为91%。它还表明估计每月节省的费用约为70万美元。此外，x-EDBOM支持跨多个供应商对相同组件进行比较分析。它将他们的价格、交货期和规格数据集成在一个统一的系统中，从而支持数据驱动的采购优化和供应商管理。除了提高效率之外，提议的框架还支持跨语言数据规范化、供应商基准测试和合并后的系统协调。它为智能产品数据管理和智能制造转型提供了可扩展的基础。

{"title":"From product specifications to smart manufacturing: Generative AI-driven process innovation for BOM digitization and its applications","authors":"Yu-Chi Lin, Jr-Fong Dang","doi":"10.1016/j.aei.2026.104451","DOIUrl":"10.1016/j.aei.2026.104451","url":null,"abstract":"<div><div>As global manufacturing advances toward digital transformation and intelligent operations, seamless integration between product specification documents and Bills of Materials (BOM) has become increasingly critical. Conventional BOM creation processes rely heavily on manual interpretation. This often leads to inconsistencies in part numbering, redundant records, and extended engineering cycles, especially when documentation is unstructured and multilingual. To address these challenges, this study proposes a generative AI–driven framework that integrates large language models (LLMs) with retrieval-augmented generation (RAG) for intelligent BOM digitization. The framework establishes a closed-loop pipeline. It converts unstructured product datasheets into rule-validated and semantically consistent part numbers through multi-stage document parsing, knowledge retrieval, and constrained decoding. A human-in-the-loop verification mechanism is incorporated to ensure semantic precision and traceability. By leveraging the multimodal capabilities of the Qwen model, the proposed framework effectively parses and understands bilingual (Chinese–English) technical documents. This enhances cross-lingual semantic comprehension and data accuracy. The validated outputs are consolidated into an executable Electrical and Digital BOM (x-EDBOM). This x-EDBOM serves as a unified source of truth across Product Lifecycle Management (PLM), Enterprise Resource Planning (ERP), and Manufacturing Execution Systems (MES). Empirical validation at a mid-sized electronics manufacturer demonstrates a 12.5× reduction in processing time (25 to 2 min), an average accuracy of 91%. It also indicates estimated monthly cost savings of approximately USD 0.7 million. Furthermore, the x-EDBOM enables comparative analysis of identical components across multiple suppliers. It integrates their price, lead-time, and specification data within a unified system, thereby supporting data-driven sourcing optimization and supplier management. Beyond efficiency improvements, the proposed framework supports cross-lingual data normalization, supplier benchmarking, and post-merger system harmonization. It provides a scalable foundation for intelligent product data management and smart manufacturing transformation.</div></div>","PeriodicalId":50941,"journal":{"name":"Advanced Engineering Informatics","volume":"71 ","pages":"Article 104451"},"PeriodicalIF":9.9,"publicationDate":"2026-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146188925","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