Jianxiong Zhang , Hongxing Qiu , Yunlong Lu , Jian Sun , Guanqi Lan
{"title":"3D reconstruction and dimension recovery algorithm for architectural structures using sequential images and photography trajectories","authors":"Jianxiong Zhang , Hongxing Qiu , Yunlong Lu , Jian Sun , Guanqi Lan","doi":"10.1016/j.autcon.2025.106465","DOIUrl":"10.1016/j.autcon.2025.106465","url":null,"abstract":"<div><div>Accurate 3D reconstruction and dimension recovery of engineering structures are crucial for dimensional inspection, whereas classic SfM + MVS reconstruction only generates scaled similar models. To address this issue, a 3D reconstruction and dimension recovery algorithm using sequential images and photography trajectories is proposed. The algorithm improves traditional 3D reconstruction by integrating inertial measurement unit (IMU) data to estimate photography trajectories, down-sampling and time-aligning trajectory points using timestamps of sequential images as reference, and solving the scale factor via data fusion to recover absolute dimensions. Laboratory validation on specimens shows that the proposed algorithm achieves a 0.58 % mean relative error (MRE) between calculated and measured dimensions, confirming high accuracy; further validation on practical structures demonstrates that, paired with the smart terminal-based mobile scheme, the proposed algorithm yields 1.29 % MRE for small-scale components and 2.73 % for large-scale structures, further verifying the engineering practicability.</div></div>","PeriodicalId":8660,"journal":{"name":"Automation in Construction","volume":"179 ","pages":"Article 106465"},"PeriodicalIF":11.5,"publicationDate":"2025-08-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144809503","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}
Yinan Xiao , Aileen Vandenberg , Dirk Lowke , Inka Mai , Pierluigi D’Acunto , Harald Kloft , Norman Hack
{"title":"Automated Robotic Assembly Planning of Space Trusses for Injection 3D Concrete Printing","authors":"Yinan Xiao , Aileen Vandenberg , Dirk Lowke , Inka Mai , Pierluigi D’Acunto , Harald Kloft , Norman Hack","doi":"10.1016/j.autcon.2025.106440","DOIUrl":"10.1016/j.autcon.2025.106440","url":null,"abstract":"<div><div>Injection 3D Concrete Printing (I3DCP) is an emerging fabrication technique that enables spatial concrete extrusion within a carrier liquid, reducing gravitational effects and allowing the creation of complex space trusses. However, I3DCP introduces new challenges in toolpath planning due to material rheology and mechanical constraints. This paper introduces an automated planning method tailored for I3DCP, integrating a constraint satisfaction problem (CSP)-based sequence planner with a Cartesian motion planner. The sequence planner uses heuristic local search with forward checking and backtracking, while the motion planner addresses end-effector redundancy with kinematic and velocity constraints. The method is validated by fabricating a 3-meter-span pedestrian bridge using a stationary 6-axis robotic arm and tested on multiple prototypes of increasing geometric complexity through simulation, demonstrating its effectiveness and scalability for intricate structural designs.</div></div>","PeriodicalId":8660,"journal":{"name":"Automation in Construction","volume":"179 ","pages":"Article 106440"},"PeriodicalIF":11.5,"publicationDate":"2025-08-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144809809","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":"Generative design for modular construction structures based on expert knowledge-informed graph neural networks and meta-heuristics","authors":"Xueqing Li, Weisheng Lu, Ziyu Peng","doi":"10.1016/j.autcon.2025.106463","DOIUrl":"10.1016/j.autcon.2025.106463","url":null,"abstract":"<div><div>With the growing demands in the construction industry, coupled with cost pressures and environmental concerns, modular building (MB) solutions are proposed to address the challenges. However, the design process of MB is more fragmented and complex, especially the structural design. This requires a reconsideration of the automated approach for its layout design, incorporating structural design. This paper develops a generative AI-enabled framework, focusing on the structural design of reinforced concrete MB. The proposed hybrid approach integrates a graph neural network-based model in a genetic generative design framework to surrogate structure design. And multiple structural related objectives are optimized in this framework. It was tested in a real project in Hong Kong and compared with the engineer's design. The optimal Pareto-balanced compromise solution resulted in a 12 % increase in usable floor area, a 7 % increase in structural performance, and a 23 % reduction in construction cost.</div></div>","PeriodicalId":8660,"journal":{"name":"Automation in Construction","volume":"179 ","pages":"Article 106463"},"PeriodicalIF":11.5,"publicationDate":"2025-08-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144810150","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":"Integrating hierarchical segmentation and vision-language reasoning for spatially complex and occluded MEP point clouds","authors":"Mingkai Li , Vincent J.L. Gan , Boyu Wang","doi":"10.1016/j.autcon.2025.106455","DOIUrl":"10.1016/j.autcon.2025.106455","url":null,"abstract":"<div><div>3D BIM reconstruction for MEP systems reduces manual documentation and enhances asset information management. However, the complexity of real-world MEP scenes, characterized by their non-linear trajectories within dense and cluttered environment, frequent data incompleteness due to occlusions, poses significant challenge for instance segmentation and geometric modeling. This paper proposes a hierarchical and progressive segmentation framework that integrates deep learning-based semantic segmentation, geometry-driven instance segmentation, and vision-language model-assisted (VLM-assisted) segmentation refinement. The semantic segmentation module achieves an overall accuracy of 87.02 % and mIoU of 69.10 %, with true positive rates exceeding 97 % for pipe, duct, and tray systems. A voxel-based DBSCAN algorithm is developed to enhance clustering stability and efficiency, followed by an improved RANSAC to extract directional primitives. In addition, VLM-assisted 2D projection analysis is introduced to refine segmentation boundaries and support downstream geometric modeling. Experimental results across multiple MEP systems demonstrate that the proposed approach achieves high segmentation accuracy and computational efficiency, without relying on large-scale annotated instance training data.</div></div>","PeriodicalId":8660,"journal":{"name":"Automation in Construction","volume":"179 ","pages":"Article 106455"},"PeriodicalIF":11.5,"publicationDate":"2025-08-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144780360","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Enhancing robustness of deep learning models against adversarial attacks for construction worker safety","authors":"Sharjeel Anjum, Chukwuma Nnaji","doi":"10.1016/j.autcon.2025.106447","DOIUrl":"10.1016/j.autcon.2025.106447","url":null,"abstract":"<div><div>The increasing use of deep neural networks (DNNs) in construction safety systems highlights their potential but also reveals vulnerabilities to adversarial perturbations. Such weaknesses can lead to false detections, increasing the risk of accidents on dynamic construction sites. This paper advances construction safety research by developing a framework to enhance AI robustness through adversarial training (AT) using the TRADES method with <span><math><msub><mi>L</mi><mo>∞</mo></msub></math></span> and <span><math><msub><mi>L</mi><mn>2</mn></msub></math></span> norms on a ResNet-18 architecture. The approach was evaluated using a combined dataset of publicly available construction images and custom-collected lab data representing unsafe behaviors. Results show the adversarially trained model achieved 92.50 % benign accuracy and 90.36 % robust accuracy under L₂ attacks. To assess model transparency, LIME (Local Interpretable Model-Agnostic Explanations) was used to visualize regions influencing predictions for both benign and adversarial inputs. These improvements support safer, AI-assisted monitoring in real-world settings by enabling more reliable decision-making and reducing the risk of AI system failures.</div></div>","PeriodicalId":8660,"journal":{"name":"Automation in Construction","volume":"179 ","pages":"Article 106447"},"PeriodicalIF":11.5,"publicationDate":"2025-08-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144780361","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}
Jinyoung Hong , Minju Kang , Hajin Choi , Shibin Lin , Heng Liu , Hoda Azari
{"title":"Automated concrete damage detection using GPR: A universal solver based on AI-assisted relative permittivity estimation","authors":"Jinyoung Hong , Minju Kang , Hajin Choi , Shibin Lin , Heng Liu , Hoda Azari","doi":"10.1016/j.autcon.2025.106453","DOIUrl":"10.1016/j.autcon.2025.106453","url":null,"abstract":"<div><div>Ground-penetrating radar (GPR) has recently been adopted for detecting damage in concrete based on relative permittivity variations. However, its practical applicability remains limited due to the need for pre-known parameters such as wave velocity or rebar depth. This paper proposes an automated algorithm that back-calculates relative permittivity from electromagnetic responses without requiring any prior structural information. Leveraging AI-assisted analysis, a YOLO-based model detects rebar-induced reflections and estimates permittivity. The algorithm was validated in three phases: (1) testing on a mock-up slab with artificial voids; (2) application to a deteriorated bridge deck, benchmarked against impact-echo results; and (3) deployment on an in-service reinforced concrete bridge. Results demonstrate high detection accuracy and significantly enhanced efficiency, enabling robust performance across varying GPR datasets. The proposed algorithm also functions as a universal solver compatible with diverse structures and equipment, advancing the automation of GPR interpretation and its broader application in civil infrastructure assessment.</div></div>","PeriodicalId":8660,"journal":{"name":"Automation in Construction","volume":"179 ","pages":"Article 106453"},"PeriodicalIF":11.5,"publicationDate":"2025-08-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144779599","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}
Yuxuan Zhang , Daniel Paes , Zhenan Feng , Dorothy Scorgie , Peijin He , Ruggiero Lovreglio
{"title":"Comparative analysis of fire evacuation decision-making in immersive vs. non-immersive virtual reality environments","authors":"Yuxuan Zhang , Daniel Paes , Zhenan Feng , Dorothy Scorgie , Peijin He , Ruggiero Lovreglio","doi":"10.1016/j.autcon.2025.106441","DOIUrl":"10.1016/j.autcon.2025.106441","url":null,"abstract":"<div><div>Understanding emergency behavior is crucial for designing safer, resilient infrastructure. Immersive Virtual Reality (VR) realistically simulates emergencies but is resource-intensive, so systematic comparisons with non-immersive VR remain scarce. To address this gap, a multifactorial VR fire-evacuation experiment was conducted in which participants navigated a room with three exits under varied conditions (e.g., social influence, smoke presence, exit distance, exit familiarity). Results indicated no significant difference in overall decision-making between immersive and non-immersive VR. Nevertheless, immersion modulated key factors: in immersive VR, participants preferred nearer exits, were more susceptible to social influence, and experienced stronger effects of smoke and exit familiarity. Smoke also reduced the influence of exit distance. Personal factors (e.g., prior VR experience, age, gender) shaped perceptions and emotions; heightened negative emotions and perceived risk were associated with less rational (i.e., suboptimal) choices, particularly in immersive VR. These insights inform VR safety training, guiding simulations that more faithfully replicate real emergencies.</div></div>","PeriodicalId":8660,"journal":{"name":"Automation in Construction","volume":"179 ","pages":"Article 106441"},"PeriodicalIF":11.5,"publicationDate":"2025-08-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144779597","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}
Cheng Yun Tsai, Mik Wanul Khosiin, Jacob J. Lin, Chuin-Shan Chen
{"title":"Multi-granular crew activity recognition for construction monitoring","authors":"Cheng Yun Tsai, Mik Wanul Khosiin, Jacob J. Lin, Chuin-Shan Chen","doi":"10.1016/j.autcon.2025.106428","DOIUrl":"10.1016/j.autcon.2025.106428","url":null,"abstract":"<div><div>The labor force is vital to construction projects, but traditional manual methods for productivity analysis are time-consuming and error-prone. Recent advancements in computer vision and deep learning offer automated solutions, yet most studies focus on low-level pose recognition, neglecting the collaborative dynamics of construction sites. This paper introduces a multi-granular crew activity recognition framework that identifies individual actions, groups collaborating workers, and links them to specific tasks. Using graph-based representations and self-attention mechanisms, the model integrates spatial and contextual information for accurate recognition. Experiments on a dataset covering rebar, formwork, and concrete operations show an overall F1 Score of 70.31%. Results highlight the importance of balancing visual features and spatial proximity for optimal performance. This framework offers an efficient solution for construction site monitoring and lays groundwork for future research on temporal modeling and human-object interaction analysis.</div></div>","PeriodicalId":8660,"journal":{"name":"Automation in Construction","volume":"179 ","pages":"Article 106428"},"PeriodicalIF":11.5,"publicationDate":"2025-08-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144779598","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}
Dongying Yang , Qing He , Honghui Wang , Yan Gao , Senhao Zhang , Guangle Yao , Meng Zhou
{"title":"High-dimensional railway vertical alignment optimization using hybrid differential evolution and gradient descent","authors":"Dongying Yang , Qing He , Honghui Wang , Yan Gao , Senhao Zhang , Guangle Yao , Meng Zhou","doi":"10.1016/j.autcon.2025.106413","DOIUrl":"10.1016/j.autcon.2025.106413","url":null,"abstract":"<div><div>Designing railway vertical alignment is challenging due to complex geometric constraints, elevation features, and cost savings expectations. Therefore, this paper proposes a Hybrid Vertical Railway Alignment Optimization (HVRAO) model to produce vertical alignment in high dimensions; the proposed model employs a parallel Differential Evolution (DE) algorithm and a swift gradient descent (GD) algorithm in turn, and a subgrade surrogate that utilizes a radial basis function is also proposed to avoid the large subgrade interpolation. Supported by a comprehensive strategy, the HVRAO model can effectively produce stable and optimal vertical alignment. Furthermore, the case study demonstrates that it is capable of creating a railway vertical alignment spanning 52 km with 56 decision variables in a matter of seconds. Finally, the proposed HVRAO framework is also applicable to horizontal alignment optimization and future bilevel alignment models in railway projects.</div></div>","PeriodicalId":8660,"journal":{"name":"Automation in Construction","volume":"179 ","pages":"Article 106413"},"PeriodicalIF":11.5,"publicationDate":"2025-08-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144773086","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}
Zhansheng Liu , Yanchi Mo , Benwei Hou , Mingming Li , Weiyi Li , Chengshun Xu
{"title":"Digital twin modelling approaches and applications in urban infrastructure operations and maintenance","authors":"Zhansheng Liu , Yanchi Mo , Benwei Hou , Mingming Li , Weiyi Li , Chengshun Xu","doi":"10.1016/j.autcon.2025.106445","DOIUrl":"10.1016/j.autcon.2025.106445","url":null,"abstract":"<div><div>Large scale and complex cascading effects make the operations and maintenance (O&M) of urban infrastructure extremely challenging. In order to identify general approaches to digital twin (DT) modelling and its applications, this review investigates 226 relevant publications as the main database. Using a systematic literature review (SLR) methodology, the database was qualitatively assessed through iterative refinements of keywords, titles, abstracts, and full texts. General procedures for DT modelling in buildings and infrastructure at different scales were summarized through quantitative analysis. To examine cascading effects in urban infrastructure O&M, DT applications were categorized into daily and abnormal scenarios and analyzed separately. Special cases, such as disasters and external disturbances, were also explored using automatic modelling and ontology fusion techniques. Finally, future research directions for DT applications in urban infrastructure are summarized to guide further development in this field.</div></div>","PeriodicalId":8660,"journal":{"name":"Automation in Construction","volume":"179 ","pages":"Article 106445"},"PeriodicalIF":11.5,"publicationDate":"2025-08-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144773087","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}