Chengzhang Chai , Yan Gao , Guanyu Xiong, Jiucai Liu, Haijiang Li
{"title":"Domain knowledge-driven image captioning for bridge damage description generation","authors":"Chengzhang Chai , Yan Gao , Guanyu Xiong, Jiucai Liu, Haijiang Li","doi":"10.1016/j.autcon.2025.106116","DOIUrl":"10.1016/j.autcon.2025.106116","url":null,"abstract":"<div><div>Deep learning-based bridge visual inspection often produces limited outputs, lacking the accurate descriptions required for practical assessments. Researchers have explored multimodal approaches to generate damage descriptions, but existing models are prone to hallucination and face challenges related to feature representation sufficiency, attention mechanism flexibility, and domain-specific knowledge integration. This paper develops an image captioning framework driven by domain knowledge to address these issues. It incorporates a multi-level feature fusion module that adaptively integrates Faster R-CNN trained weights (domain knowledge) with a CNN architecture. Additionally, it introduces a correlation-aware attention mechanism to dynamically capture interdependencies between image regions and optimise the attentional focus during LSTM decoding. Experimental results show that the proposed framework achieves higher BLEU scores and improves image-text alignment as verified through attention heatmaps. While the framework enhances inspection efficiency and quality, further dataset expansion and broader domain validation are required to assess its generalisation ability.</div></div>","PeriodicalId":8660,"journal":{"name":"Automation in Construction","volume":"174 ","pages":"Article 106116"},"PeriodicalIF":9.6,"publicationDate":"2025-03-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143621397","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}
Zhenqiang Han , Jiaqi Tang , Liqun Hu , Wei Jiang , Aimin Sha
{"title":"Automated measurement of asphalt pavement rut depth using smartphone imaging","authors":"Zhenqiang Han , Jiaqi Tang , Liqun Hu , Wei Jiang , Aimin Sha","doi":"10.1016/j.autcon.2025.106124","DOIUrl":"10.1016/j.autcon.2025.106124","url":null,"abstract":"<div><div>Fast and accurate rutting distress detection is essential for driving safety and advancing pavement maintenance automation. However, existing Rut Depth (RD) measurement methods are often inefficient or costly due to complex pavement conditions and expensive equipment. This paper introduces a method to identify and measure RD using smartphone photography and neural networks. Accelerated Pavement Tests (APTs) were conducted to capture rutting evolution, combining measured and photographed data. Grayscale images were analyzed via Fourier transform, identifying the rut-related frequency range (0–0.02 Hz). Grayscale rut curves corresponding to actual rut cross-sections were extracted, and seven feature points were used to calculate grayscale RD. A backpropagation neural network model was trained and validated, demonstrating RD detection within 5–45 mm, with an average absolute error of 1.29 mm. This method provides an alternative for efficient RD measurement in APT and field pavement condition evaluations, offering potentials for improving pavement maintenance practices.</div></div>","PeriodicalId":8660,"journal":{"name":"Automation in Construction","volume":"174 ","pages":"Article 106124"},"PeriodicalIF":9.6,"publicationDate":"2025-03-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143609936","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}
Yi Zhou , Bowen Huang , Boge Dong , Yi Wen , Molong Duan
{"title":"Dynamic robotic bricklaying force-position control considering mortar dynamics for enhanced consistency","authors":"Yi Zhou , Bowen Huang , Boge Dong , Yi Wen , Molong Duan","doi":"10.1016/j.autcon.2025.106090","DOIUrl":"10.1016/j.autcon.2025.106090","url":null,"abstract":"<div><div>With growing construction automation needs and aging workforce, robotic bricklaying technologies offer promising solutions by replacing labor-intensive manual wall construction. Despite advances in mechatronics, trajectory generation, and adhesive bonding, bricklaying process consistency still remains a major issue, challenged by nonlinear mortar dynamics, variable thickness, and lack of force-position control. This paper enhances consistency by adjusting the robot trajectory to achieve the desirable maximum contact force and brick placement accuracy. The feedforward bricklaying trajectory is generated by solving an optimal control problem with unspecified terminal time to approach the desirable maximum force while maintaining placement accuracy, considering the identified mortar dynamics. To ensure precise positioning and robust desired maximum contact force, robotic bricklaying force-position feedback control is proposed to dynamically control the velocity utilizing real-time force feedback. Simulations and experiments validated the proposed approach, including shear tests, demonstrating enhanced consistency and the relationship between maximum contact force and bond strength.</div></div>","PeriodicalId":8660,"journal":{"name":"Automation in Construction","volume":"174 ","pages":"Article 106090"},"PeriodicalIF":9.6,"publicationDate":"2025-03-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143610036","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}
Yuntae Jeon , Dai Quoc Tran , Khoa Tran Dang Vo , Jaehyun Jeon , Minsoo Park , Seunghee Park
{"title":"Corrigendum to “Neural radiance fields for construction site scene representation and progress evaluation with BIM” [Automation in Construction, Volume 172 (2025) 106013]","authors":"Yuntae Jeon , Dai Quoc Tran , Khoa Tran Dang Vo , Jaehyun Jeon , Minsoo Park , Seunghee Park","doi":"10.1016/j.autcon.2025.106121","DOIUrl":"10.1016/j.autcon.2025.106121","url":null,"abstract":"","PeriodicalId":8660,"journal":{"name":"Automation in Construction","volume":"173 ","pages":"Article 106121"},"PeriodicalIF":9.6,"publicationDate":"2025-03-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143738292","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Mingqiao Han, Jihan Zhang, Yijun Huang, Jiwen Xu, Xi Chen, Ben M. Chen
{"title":"Enhancing worker monitoring and management on large-scale construction sites with UAVs and digital twin modeling","authors":"Mingqiao Han, Jihan Zhang, Yijun Huang, Jiwen Xu, Xi Chen, Ben M. Chen","doi":"10.1016/j.autcon.2025.106108","DOIUrl":"10.1016/j.autcon.2025.106108","url":null,"abstract":"<div><div>Monitoring large-scale work sites is challenging, particularly in vast outdoor areas. Unmanned aerial vehicles (UAVs) provide an effective solution for site monitoring and worker management. This paper introduces a UAV-based framework integrated with digital twin (DT) modeling to enhance real-time data management and worker authorization verification. The pretrained YOLO-LCA model improved detection accuracy from 31.5% to 96.4%. The framework combines multi-object tracking with 3D site reconstruction, enabling precise global registration and situational awareness. Cross-referencing UAV detections with GPS-enabled worker IDs ensures that only authorized personnel are present, effectively identifying unapproved workers. The proposed framework has undergone large-scale validation across multiple construction projects in Hong Kong, demonstrating significant potential for modernizing work site management. By integrating UAVs and DT technology, this framework supports efficient monitoring, operational safety, and informed decision-making, providing a scalable approach to addressing the demands of large-scale construction site management.</div></div>","PeriodicalId":8660,"journal":{"name":"Automation in Construction","volume":"174 ","pages":"Article 106108"},"PeriodicalIF":9.6,"publicationDate":"2025-03-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143601474","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}
Yudong Liu, Han Huang, Ge Gao, Ziyi Ke, Shengtao Li, Ming Gu
{"title":"Dataset and benchmark for as-built BIM reconstruction from real-world point cloud","authors":"Yudong Liu, Han Huang, Ge Gao, Ziyi Ke, Shengtao Li, Ming Gu","doi":"10.1016/j.autcon.2025.106096","DOIUrl":"10.1016/j.autcon.2025.106096","url":null,"abstract":"<div><div>As-built BIM reconstruction plays a significant role in urban renewal and building digitization but currently faces challenges of low efficiency. Scan-to-BIM aims to improve reconstruction efficiency but lacks domain-specific, large-scale datasets and accurate, multi-dimensional benchmark metrics. These deficiencies further impede the evaluation and training of scan-to-BIM methods. To address these challenges, this paper proposes BIMNet, an IFC-based large-scale point cloud to BIM dataset, and a set of metrics that reflect the quality and issues of reconstructed models from both geometric and topological perspectives. Experiments demonstrate that BIMNet enhances the evaluation and training of scan-to-BIM methods during the critical processes of reconstruction and segmentation. This research contributes to the data foundation and metric system for deep-learning based scan-to-BIM methods. In the future, BIMNet will not only facilitate the development of scan-to-BIM but also contribute to the advancement of smart cities and AI-driven technologies beyond scan-to-BIM.</div></div>","PeriodicalId":8660,"journal":{"name":"Automation in Construction","volume":"173 ","pages":"Article 106096"},"PeriodicalIF":9.6,"publicationDate":"2025-03-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143591537","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}
Yuhan Li , Xiaopeng Yang , Junbo Gong , Jian Wang , Zihang Jiang , Tian Lan
{"title":"Intelligent detection of bonding status in external building insulation layers using ground-penetrating radar","authors":"Yuhan Li , Xiaopeng Yang , Junbo Gong , Jian Wang , Zihang Jiang , Tian Lan","doi":"10.1016/j.autcon.2025.106100","DOIUrl":"10.1016/j.autcon.2025.106100","url":null,"abstract":"<div><div>The bonding status of external building insulation layer is crucial for thermal insulation and long-term safety, but existing detection methods lack efficiency and accuracy. This paper explores the use of Ground-Penetrating Radar (GPR) technology for accurately estimating bonding areas and precisely identifying top and subgrade debonding defects in external building insulation layers. It proposes an end-to-end intelligent detection method based on GPR, incorporating a multi-task branch network that automatically selects C-scan depth slices for semantic segmentation to estimate bonding areas and utilizes B-scan slices for target detection of debonding defects. Results show that area estimation’s relative error is 0.70%, debonding detection accuracy reaches 78.45%, and the model performs well in complex scenarios. This paper provides the application of GPR in building inspection, promoting hazard discovery and technological advancement. Future work will focus on improving clutter suppression algorithms and C-scan depth slice extraction methods to further enhance detection results.</div></div>","PeriodicalId":8660,"journal":{"name":"Automation in Construction","volume":"173 ","pages":"Article 106100"},"PeriodicalIF":9.6,"publicationDate":"2025-03-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143591535","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":"Integrated algorithm for identifying failure modes and assessing reliability of concrete-filled steel tubes under lateral impact","authors":"Nan Xu, Yanhui Liu","doi":"10.1016/j.autcon.2025.106118","DOIUrl":"10.1016/j.autcon.2025.106118","url":null,"abstract":"<div><div>Concrete-filled steel tube (CFST) columns are susceptible to transverse impact and catastrophic fracture failure might trigger progressive collapse of entire buildings. This paper aims to predict CFST failure modes (bending deformation, crack and fracture) and conduct reliability evaluation implementing intelligent algorithms. Fixed-supported CFST impact samples were gathered, which contain 10 inputs and 3 outputs (crack deflection, fracture deflection and maximum deflection). Results indicated that support vector regression (SVR) predicted three outputs optimally adopting RBF kernel function, and osprey optimization algorithm (OOA) optimizing SVR achieved more superior prediction than particle swarm optimization. Three output variables were utilized to identify CFST failure modes, OOA-SVR(R) possessed 97.70 % accuracy for 87 test samples. Monte Carlo sampling was conducted to estimate fracture vulnerability curves considering random distribution of input variables. Finally, a performance-based CFST design procedure was proposed, declining computational requirements and strengthening user-friendliness in contrast to conventional approaches.</div></div>","PeriodicalId":8660,"journal":{"name":"Automation in Construction","volume":"173 ","pages":"Article 106118"},"PeriodicalIF":9.6,"publicationDate":"2025-03-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143591536","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}
Zihao Jin , Wei Zhang , Zhenyu Yin , Ning Zhang , Xueyu Geng
{"title":"Estimating track geometry irregularities from in-service train accelerations using deep learning","authors":"Zihao Jin , Wei Zhang , Zhenyu Yin , Ning Zhang , Xueyu Geng","doi":"10.1016/j.autcon.2025.106114","DOIUrl":"10.1016/j.autcon.2025.106114","url":null,"abstract":"<div><div>Timely identification of Track Geometry Irregularities (TGIs) is essential for ensuring the safety and comfort of high-speed rail operations. Existing inspection methods rely on costly Track Recording Vehicles (TRVs) and manual trolleys, resulting in infrequent and expensive inspections. This paper proposes a data-driven approach for estimating TGIs using a Convolutional Neural Network with Multi-Head and Multi-Layer Perceptron (CNN-MH-MLP) architecture. A comprehensive vehicle-track-embankment-ground Finite Element (FE) model incorporating geometric wheel-rail nonlinearity is developed to generate the in-service train acceleration data used for training the network. The CNN-MH-MLP network demonstrates strong performance in estimating TGIs, exhibiting robustness to noise. Optimized sensor placement with three sensors achieves the best trade-off between accuracy and efficiency. Furthermore, the network's transferability highlights the significance of detailed numerical models in producing virtual databases. This work is expected to facilitate the development of intelligent systems for real-time TGI monitoring, improving inspection efficiency and reducing labor costs.</div></div>","PeriodicalId":8660,"journal":{"name":"Automation in Construction","volume":"173 ","pages":"Article 106114"},"PeriodicalIF":9.6,"publicationDate":"2025-03-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143579210","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Large language model-empowered paradigm for automated geotechnical site planning and geological characterization","authors":"Zehang Qian, Chao Shi","doi":"10.1016/j.autcon.2025.106103","DOIUrl":"10.1016/j.autcon.2025.106103","url":null,"abstract":"<div><div>A sound site investigation scheme must satisfy requirements of various local, regional, or national codes, and it is imperative to have an efficient system for information retrieval, summarization, and reasoning along with a rapid interpretation tool for real-time risk-informed decision-making. Emerging large language models (LLMs) offer a promising solution for automatically processing unstructured natural languages and facilitating collaborative reasoning between humans and machines. This paper develops a customized LLM-based agent named “Geologist” to streamline geotechnical site planning and subsequent geological interpretation. A Multihop-Retrieval-Augmented Generation system is proposed to retrieve site-specific requirements from multiple site investigation design codes. Moreover, a progressive human-machine collaboration scheme is orchestrated for interpretable geological modelling. The efficiency of the proposed LLM-guided paradigm is validated through illustrative examples and real-world case histories. Results show that the proposed workflow facilitates real-time and accurate information retrieval as well as automatic development of subsurface geological cross-sections with quantified stratigraphic uncertainty.</div></div>","PeriodicalId":8660,"journal":{"name":"Automation in Construction","volume":"173 ","pages":"Article 106103"},"PeriodicalIF":9.6,"publicationDate":"2025-03-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143579209","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}