{"title":"Generative design for early-stage optioneering in subsea layout design","authors":"Bilal Muhammed, Soban Babu Beemaraj, Amol Joshi","doi":"10.1016/j.autcon.2025.106561","DOIUrl":"https://doi.org/10.1016/j.autcon.2025.106561","url":null,"abstract":"The conceptual stage of large infrastructure construction projects, such as subsea layout design, is time-consuming, costly, and reliant on significant human effort. Generative Design (GD) is a powerful tool for decision support, reducing design timelines, and improving the quality of decision-making through the rapid automatic exploration of large design spaces. In this paper, a generalized GD approach is introduced for flow network design, which is then applied to subsea layout design. Various algorithmic strategies address practical challenges and constraints, yielding feasible design options for subproblems, including drill center clustering, manifold positioning, process host positioning, and flowline design. The proposed systematic design process flow facilitates the iterative refinement of layout options based on designer preferences, promoting adaptive decision-making. When applied to a real-world oil and gas field in Angola, this GD approach generates several feasible subsea layouts ranked by total development cost, showcasing its potential for cost-effective and safer designs.","PeriodicalId":8660,"journal":{"name":"Automation in Construction","volume":"557 1","pages":""},"PeriodicalIF":10.3,"publicationDate":"2025-09-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145181250","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Visual question answering for intelligent maintenance of maglev railway systems","authors":"Gao-Feng Jiang , Su-Mei Wang , Yi-Qing Ni","doi":"10.1016/j.autcon.2025.106559","DOIUrl":"10.1016/j.autcon.2025.106559","url":null,"abstract":"<div><div>Traditional methodologies effectively assess the railway components by categorizing them as either normal or abnormal. However, these methods provide limited insight into underlying conditions and the reasons for abnormalities, often lacking a comprehensive explanation. With advancements in multimodal feature learning, multimodal data are potentially integrated for various downstream tasks, such as visual question answering (VQA). This paper proposes a three-phase procedure for VQA-guided maintenance of maglev conditions, aiming to automatically recognize evidence of damage and failure using accumulated multimodal knowledge. As one of the early VQA datasets designed for railway condition monitoring, it is organized as image-question-answer tuples, where images are generated from time-frequency spectrograms, questions and answers are formulated based on maglev structural dynamic characteristics. The results indicate that the proposed model is reliable in answer accuracy and expression quality. This advancement contributes to forming intelligent decision-making processes in railway infrastructure, ultimately promoting safer and more efficient railway operations.</div></div>","PeriodicalId":8660,"journal":{"name":"Automation in Construction","volume":"180 ","pages":"Article 106559"},"PeriodicalIF":11.5,"publicationDate":"2025-09-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145155679","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}
Wenli Liu , Zhaoji Wu , Xingyu Tao , Helen H.L. Kwok , Jack C.P. Cheng
{"title":"Automated simulation modeling and sustainable performance evaluation for flat layout design","authors":"Wenli Liu , Zhaoji Wu , Xingyu Tao , Helen H.L. Kwok , Jack C.P. Cheng","doi":"10.1016/j.autcon.2025.106556","DOIUrl":"10.1016/j.autcon.2025.106556","url":null,"abstract":"<div><div>Generative design is increasingly applied to flat layouts. However, generative models lack automated and efficient methods for performance evaluation, including (1) automatic conversion of flat layout images into simulation models and (2) efficient performance evaluation at the flat level. This paper addresses these issues by proposing an automated image-based performance evaluation surrogate model. Firstly, a new geometric feature set is proposed. Secondly, the Image2Sim algorithm is developed for automated flat layout modeling from images in the RPLAN dataset. Finally, a graph-aware extreme gradient boosting (GAXGBoost) surrogate model is developed for flat-level performance evaluation. Results demonstrate that (1) the Image2Sim algorithm reduces the failure rate by 8 % and simulation modeling time from 333 days to 2 days; (2) the GAXGBoost outperforms XGBoost, MARS, GNN, and ANN across all metrics. The GAXGBoost provides accurate and timely feedback on flat layout performance, thus facilitating performance-driven generative design in the early design stage.</div></div>","PeriodicalId":8660,"journal":{"name":"Automation in Construction","volume":"180 ","pages":"Article 106556"},"PeriodicalIF":11.5,"publicationDate":"2025-09-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145155726","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}
Yaxian Dong , Zijun Zhan , Yuqing Hu , Daniel Mawunyo Doe , Zhu Han
{"title":"AI BIM coordinator for non-expert interaction in building design using LLM-driven multi-agent systems","authors":"Yaxian Dong , Zijun Zhan , Yuqing Hu , Daniel Mawunyo Doe , Zhu Han","doi":"10.1016/j.autcon.2025.106563","DOIUrl":"10.1016/j.autcon.2025.106563","url":null,"abstract":"<div><div>The building construction domain relies heavily on human experts for specialized tasks, often leading to excessive workload and inefficiency. For example, building design coordination typically requires BIM coordinators to operate BIM tools on participants' behalf, with limited ability to capture procedural knowledge. To address this gap, this paper introduces AI-based BIM coordinators to enable non-BIM experts to seamlessly interact with BIM tools while automatically documenting valuable procedural information. The approach customizes five LLM-based agents equipped with fundamental building knowledge and skills. Through user prompts, the collaboration of “bim_assistant” and “checker” agents ensures the accuracy and consistency of generated code, while the “sender” and “executor” agents facilitate direct execution in BIM tools beyond text-based responses, with the “terminator” ending the process. An AutoGen-Revit prototype demonstrates feasibility, flexibility, robustness, and time-cost effectiveness for common BIM-based design coordination tasks, including space reasoning, 3D element creation, clash detection, informative 3D view saving, documentation, and retrieval. This paper enables specialist-free workflows and generates procedural logs and images that serve as datasets for future analysis.</div></div>","PeriodicalId":8660,"journal":{"name":"Automation in Construction","volume":"180 ","pages":"Article 106563"},"PeriodicalIF":11.5,"publicationDate":"2025-09-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145155677","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Rapid seismic loss assessment of school buildings using adaptive basis vectors sampling for support vector regression","authors":"Wenkai Shi , Huan Luo","doi":"10.1016/j.autcon.2025.106547","DOIUrl":"10.1016/j.autcon.2025.106547","url":null,"abstract":"<div><div>Accurate and rapid assessment of earthquake-induced economic losses across varying seismic intensities and the exploration of their relationship are crucial for quantifying structural seismic resilience. Traditional seismic loss assessments using the finite element method require computationally intensive nonlinear time-history analyses, while conventional machine learning methods demand substantial resources for training on large seismic response datasets. This paper proposes Adaptive Basis Vectors Sampling Guided Support Vector Machines for Regression (ABVS-SVMR) to overcome these limitations. ABVS-SVMR reduces computational complexity by adaptively sampling an optimal subset to construct a low-rank kernel matrix approximating the full matrix, improving training efficiency. Bench-marking against Least Squares SVMR (LS-SVMR) and Artificial Neural Networks (ANNs) using 18,438 seismic loss records from reinforced concrete frame school buildings demonstrated excellent predictive accuracy (<span><math><mrow><msup><mrow><mi>R</mi></mrow><mrow><mn>2</mn></mrow></msup><mo>></mo><mn>0</mn><mo>.</mo><mn>97</mn></mrow></math></span>) for all models. Crucially, ABVS-SVMR achieved a 13-fold speedup over LS-SVMR and a 27-fold speedup over ANNs, demonstrating exceptional potential for rapid seismic loss assessment.</div></div>","PeriodicalId":8660,"journal":{"name":"Automation in Construction","volume":"180 ","pages":"Article 106547"},"PeriodicalIF":11.5,"publicationDate":"2025-09-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145155678","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}
Zhongchen Deng , Zhechen Yang , Chi Chen , Cheng Zeng , Yan Meng , Bisheng Yang
{"title":"Multimodal plane instance segmentation with the Segment Anything Model","authors":"Zhongchen Deng , Zhechen Yang , Chi Chen , Cheng Zeng , Yan Meng , Bisheng Yang","doi":"10.1016/j.autcon.2025.106541","DOIUrl":"10.1016/j.autcon.2025.106541","url":null,"abstract":"<div><div>Plane instance segmentation from RGB-D data is critical for BIM-related tasks. However, existing deep-learning methods rely on only RGB bands, overlooking depth information. To address this, PlaneSAM, a Segment-Anything-Model-based network, is proposed. It fully integrates RGB-D bands using a dual-complexity backbone: a simple branch primarily for the D band and a high-capacity branch mainly for RGB bands. This structure facilitates effective D-band learning with limited data, preserves EfficientSAM’s RGB feature representations, and enables task-specific fine-tuning. To improve adaptability to RGB-D domains, a self-supervised pretraining strategy is introduced. EfficientSAM’s loss is also optimized for large-plane segmentation. Additionally, plane detection is performed using Faster R-CNN, enabling fully automatic segmentation. State-of-the-art performance is achieved on multiple datasets, with <span><math><mo><</mo></math></span>10% additional overhead compared to EfficientSAM. The proposed dual-complexity backbone shows strong potential for transferring RGB-based foundation models to RGB<span><math><mo>+</mo></math></span>X domains in other scenarios, while the pretraining strategy is promising for other data-scarce tasks.</div></div>","PeriodicalId":8660,"journal":{"name":"Automation in Construction","volume":"180 ","pages":"Article 106541"},"PeriodicalIF":11.5,"publicationDate":"2025-09-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145155727","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Robust three-stage deep learning and image processing framework for automated loose bolt detection in complex environments","authors":"Yaqi Wang , Xiukun Wei , Donghua Wu , Siqi Wu , Huaze Xia","doi":"10.1016/j.autcon.2025.106531","DOIUrl":"10.1016/j.autcon.2025.106531","url":null,"abstract":"<div><div>Vision-based bolt looseness detection is critical for infrastructure safety, yet current methods struggle with bolts of diverse scales, types, and viewing angles in complex environments. This research addresses the challenge of achieving accurate looseness identification for multi-type bolts under such conditions. A three-stage framework is proposed that decouples the task into bolt localization using improved YOLOv8, fine-grained classification via the lightweight RepViT network, and multi-strategy looseness recognition of image processing and deep learning. The method achieves high accuracy and efficiency across all stages, with localization recall at 96.1%, classification accuracy at 98.4%, and final looseness identification accuracy up to 94.5%. This research will advance the application of machine vision in defect identification and intelligent maintenance within the construction sector. The phased methodology may similarly be applied to defect detection in other infrastructure domains, and extended to develop end-to-end integrated systems.</div></div>","PeriodicalId":8660,"journal":{"name":"Automation in Construction","volume":"180 ","pages":"Article 106531"},"PeriodicalIF":11.5,"publicationDate":"2025-09-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145155675","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}
Xiaochun Luo , Yutong Tang , Yongqi Wei , Chengqian Li , Qi Fang
{"title":"GNN-based spatial relationship modeling for automated scaffold component function recognition and intelligent compliance checking","authors":"Xiaochun Luo , Yutong Tang , Yongqi Wei , Chengqian Li , Qi Fang","doi":"10.1016/j.autcon.2025.106557","DOIUrl":"10.1016/j.autcon.2025.106557","url":null,"abstract":"<div><div>Manual scaffold inspection is inefficient and error-prone, particularly for complex and large-scale structures. Existing scan-to-BIM methods rely on hard-coded rules and the results lack sufficient semantic richness, limiting automation and scalability for comprehensive compliance checking. This paper presents an approach to integrating images and point clouds for automating scaffold component function recognition and compliance checking. Two scaffold graph designs—Tube Node Graph (TNG) and Tube-Plane Node Graph (TPNG)—are proposed, employing Graph Neural Networks (GNNs) to model spatial relationships and identify scaffold tube member functions. The primary distinction between TNG and TPNG is whether wall and ground plane elements are represented as nodes in the graph. Evaluation results show that TPNG outperforms TNG, achieving recognition accuracies of 84.86 % and 73.03 %, respectively. The proposed method enhances the efficiency and accuracy of scaffold compliance checking, providing an effective solution for automated inspection.</div></div>","PeriodicalId":8660,"journal":{"name":"Automation in Construction","volume":"180 ","pages":"Article 106557"},"PeriodicalIF":11.5,"publicationDate":"2025-09-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145155676","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}
Lizhi Long , Wenyao Liu , Shaopeng Xu , Peng Shi , Cheng Zhang , Lu Deng
{"title":"Automated alignment deviation measurement for precast concrete assembly using point cloud-image fusion","authors":"Lizhi Long , Wenyao Liu , Shaopeng Xu , Peng Shi , Cheng Zhang , Lu Deng","doi":"10.1016/j.autcon.2025.106540","DOIUrl":"10.1016/j.autcon.2025.106540","url":null,"abstract":"<div><div>Current Precast concrete (PC) column assembly methods face difficulties in precisely measuring the deviation between rebars and sleeves. This paper proposes an automated assembly alignment deviation measurement (AADM) method that integrates 3D point cloud data with 2D images through complementary algorithms. The proposed method comprises a Virtual Trial Assembly (VTA) module that extract sleeve and rebar assembly points and an Alignment Deviation Measurement (ADM) module that calculate rebar-sleeve deviation using three non-collinear points extracted from VTA module. Deviation measurement experiments were conducted on both a PC column model and a real component. Results show that the proposed AADM method outperforms the investigated benchmark methods, with extraction errors lower than 1 mm and positioning accuracy within 3 mm, both meeting the specification requirements. These findings indicate the proposed method enables precise deviation measurement before and during hoisting, providing assistance for automated alignment of precast concrete columns.</div></div>","PeriodicalId":8660,"journal":{"name":"Automation in Construction","volume":"180 ","pages":"Article 106540"},"PeriodicalIF":11.5,"publicationDate":"2025-09-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145155674","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Explainable artificial intelligence (XAI)-driven probabilistic image-based structural health monitoring of reinforced concrete beams with shear reinforcements","authors":"Qisen Chen , Bing Li","doi":"10.1016/j.autcon.2025.106549","DOIUrl":"10.1016/j.autcon.2025.106549","url":null,"abstract":"<div><div>This paper addresses the challenge of accurately evaluating structural damage in reinforced concrete (RC) beams with shear reinforcements using image-based data. The research question focuses on whether probabilistic and explainable machine learning models can effectively predict strength- and displacement-based damage indicators from crack images. A framework is developed that integrates Explainable Artificial Intelligence (XAI), probabilistic shear strength modeling, image processing, and feature selection to extract 41 critical damage-related features. Four machine learning models are trained and validated using 375 images from ten experimental studies, with Gaussian Process Regression achieving an R<sup>2</sup> value of 0.923 in strength-based prediction. These results offer a non-contact, scalable, and interpretable solution for structural health monitoring and safety assessment of RC members. The findings encourage further exploration of image-based and probabilistic SHM approaches under cyclic, seismic, or environmental loading conditions.</div></div>","PeriodicalId":8660,"journal":{"name":"Automation in Construction","volume":"180 ","pages":"Article 106549"},"PeriodicalIF":11.5,"publicationDate":"2025-09-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145118968","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}