Ensieh Ali Bakhshi , Omid Yazdanpanah , Kiarash M. Dolatshahi
{"title":"Pixel-level multicategory semantic segmentation of visible seismic damage in bridge piers using an Attention-Mamba Transformer-based U-Net model","authors":"Ensieh Ali Bakhshi , Omid Yazdanpanah , Kiarash M. Dolatshahi","doi":"10.1016/j.autcon.2025.106587","DOIUrl":"10.1016/j.autcon.2025.106587","url":null,"abstract":"<div><div>Currently, many computer vision-based studies focus on cyclic test photos approximating structural behavior under seismic loads and struggling with severely imbalanced multiclass seismic damage detection, particularly cracks. This paper presents an approach for pixel-level detection of visible seismic damage in RC bridge piers, identifying cracks, spalling, reinforcement exposure, crushing, and buckling/failure. A semantic segmentation database is built from experimental images emphasizing real-time hybrid simulations, with lens correction, perspective adjustment, and augmentation to enhance diversity. Hypergeometric distribution and weighted loss functions address class imbalance at both sample and pixel levels. A self-attention-Mamba-driven transformer block with inception modules is integrated into a customized U-Net bottleneck, achieving per-class IoU over 0.7958. A VGG16 encoder with Mamba blocks further refines crack feature extraction (length, width, angles), reaching IoU above 0.6478. Overlapping and mirror padding improve mask blending. The model generalizes well to unseen bridge piers and shear walls, supporting accurate post-earthquake damage assessment.</div></div>","PeriodicalId":8660,"journal":{"name":"Automation in Construction","volume":"181 ","pages":"Article 106587"},"PeriodicalIF":11.5,"publicationDate":"2025-10-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145314974","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}
Ankang Ji , Limao Zhang , Yudan Dou , Yuexiong Ding , Minggong Zhang , Luqi Wang
{"title":"Dual heterogeneous attention-based deep learning model for multi-output prediction of TBM operations","authors":"Ankang Ji , Limao Zhang , Yudan Dou , Yuexiong Ding , Minggong Zhang , Luqi Wang","doi":"10.1016/j.autcon.2025.106605","DOIUrl":"10.1016/j.autcon.2025.106605","url":null,"abstract":"<div><div>Predicting tunnel boring machine (TBM) performance in real-time is challenging due to the complex, dynamic, and multi-output nature of TBM operations. To address the challenges, this paper proposes a deep-learning method to provide an effective and efficient solution for predicting multi-output TBM performance in real-time, while also guiding TBM operations. This method integrates various essential components, including two parallel bi-directional long short-term memory (BiLSTM), a dual heterogeneous attention module (DHAM), a loss function, and evaluation metrics to ensure precise predictions while maintaining computational efficiency for real-time deployment. Experiments on real-world TBM operation data showcase the model's enhanced capabilities, achieved through the model featuring the learning rate of 0.00001, the batch size of 4, the full training set, the 2-step time window, the utilizations of the Nadam optimizer and the DHAM, and the ensemble of multiple modules. A comparative analysis reveals that the proposed method outperforms existing state-of-the-art models. This paper not only demonstrates the capabilities of the proposed method but also opens up opportunities for further advancements in utilizing deep learning to enhance decision-making processes and operational efficiency within the infrastructure construction fields.</div></div>","PeriodicalId":8660,"journal":{"name":"Automation in Construction","volume":"181 ","pages":"Article 106605"},"PeriodicalIF":11.5,"publicationDate":"2025-10-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145314973","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":"Fully automated synthetic BIM dataset generation using a deep learning-based framework","authors":"Xing Liang , Nobuyoshi Yabuki , Tomohiro Fukuda","doi":"10.1016/j.autcon.2025.106584","DOIUrl":"10.1016/j.autcon.2025.106584","url":null,"abstract":"<div><div>Building information models (BIMs) are essential for efficient building operation, yet most existing buildings only have two-dimensional (2D) drawings, leading to increased interest in 2D-to-BIM reconstruction. To address the data scarcity hindering automated BIM reconstruction and evaluation, this paper presents a deep learning-based fully automated framework for BIM dataset generation. The approach uses image processing to define polygonal boundaries, applies neural networks to generate geometric layouts, and augments semantic information with predefined data for BIM generation via software application programming interfaces (APIs). The resulting Residential unit BIM (ResBIM) is a synthetic dataset comprising over 1000 paired BIMs (RVT format) and their corresponding 2D floor plans automatically annotated via a toolbox, filling a critical gap in BIM data availability. This work provides a scalable automated BIM reconstruction solution and establishes the foundation for future AI-driven BIM automation research.</div></div>","PeriodicalId":8660,"journal":{"name":"Automation in Construction","volume":"181 ","pages":"Article 106584"},"PeriodicalIF":11.5,"publicationDate":"2025-10-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145314975","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}
Nasrullah Khan , Dohyeong Kim , Minju Kim , Daeho Kim , Dongmin Lee
{"title":"Message-passing framework for multi-camera worker tracking in construction","authors":"Nasrullah Khan , Dohyeong Kim , Minju Kim , Daeho Kim , Dongmin Lee","doi":"10.1016/j.autcon.2025.106610","DOIUrl":"10.1016/j.autcon.2025.106610","url":null,"abstract":"<div><div>Current computer vision–based tracking systems face challenges in reliably associating identities of construction workers due to similar attire, frequent occlusions, and complex multi-view movements, leading to fragmented trajectories and ID switches. This work proposes a multi-camera tracking framework that detects workers in individual camera views and integrates observations across cameras using re-identification and message-passing. A region-based re-identification model enhances feature extraction for occluded workers and those wearing similar gear, producing more discriminative representations. Data association leverages message-passing approach to combine localization, visual features, and motion cues for robust clustering and trajectory generation. Experiments achieve IDF1 scores of 68.30 (controlled) and 85.10 (outdoor) with MOTA scores of 79.7 and 79.2, respectively. Results on the CAMPUS benchmark demonstrate strong generalization and competitive performance, meeting operational requirements for field deployment. The approach highlights potential for broader multi-camera tracking in occluded industrial environments, supporting applications in safety and productivity monitoring.</div></div>","PeriodicalId":8660,"journal":{"name":"Automation in Construction","volume":"181 ","pages":"Article 106610"},"PeriodicalIF":11.5,"publicationDate":"2025-10-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145314976","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}
Ignasi Fernandez , Carlos G. Berrocal , Mikael Johansson , Mattias Roupe , Rasmus Rempling
{"title":"Enhanced digital twin for on-site inspections using distributed optical fiber sensors and augmented reality","authors":"Ignasi Fernandez , Carlos G. Berrocal , Mikael Johansson , Mattias Roupe , Rasmus Rempling","doi":"10.1016/j.autcon.2025.106602","DOIUrl":"10.1016/j.autcon.2025.106602","url":null,"abstract":"<div><div>Infrastructure inspections are still largely manual, episodic, and subjective, which delays damage detection and limits data-informed decision making. The paper introduces a Digital Twin framework designed to enhance infrastructure inspections using Distributed Optical Fiber Sensors (DOFS) and Augmented Reality (AR). The framework integrates advanced sensing technologies, edge computing, and web-based applications to provide real-time and historical data visualization during inspections. DOFS technology, known for its high spatial resolution and sensitivity to strain and temperature variations, is utilized to capture high-resolution strain data for continuous structural health monitoring. The framework combines DOFS data with Building Information Modelling (BIM) and AR to create a virtual representation of the assets, enabling precise and efficient on-site inspections. Two case studies demonstrate the practical application of this system: one focusing on historical data visualization and the other on real-time sensor data visualization. The results highlight the framework's ability to provide valuable insights into infrastructure health, improve inspection accuracy, and enhance decision-making processes.</div></div>","PeriodicalId":8660,"journal":{"name":"Automation in Construction","volume":"181 ","pages":"Article 106602"},"PeriodicalIF":11.5,"publicationDate":"2025-10-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145314977","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}
Jinze Luo , Qianxun Yang , Yumeng Sun , Lunpeng Li , Wenyuan Cai , Yuchuan Du , Chenglong Liu
{"title":"Quadruped robot-enabled framework for intelligent sidewalk condition monitoring","authors":"Jinze Luo , Qianxun Yang , Yumeng Sun , Lunpeng Li , Wenyuan Cai , Yuchuan Du , Chenglong Liu","doi":"10.1016/j.autcon.2025.106600","DOIUrl":"10.1016/j.autcon.2025.106600","url":null,"abstract":"<div><div>Accurate detection of sidewalk conditions is critical for enhancing the safety and efficiency of urban non-motorized transportation systems. Existing manual inspection methods, while effective for limited scenarios, fall short in terms of efficiency, coverage, and the detection of concealed defects. This paper introduces a sidewalk inspection framework that leverages a quadruped robot equipped with visual and vibration sensors to comprehensively address complex pavement scenes and hidden loosened bricks. For visual inspection, an improved YOLO11 model with a dynamic attention-weighted sampling mechanism enhances detection accuracy for minority-class defects, achieving a mAP of 0.883. For hidden defect identification, a Physics-Informed Neural Network with a Long Short-Term Memory network (PINN-LSTM) is proposed to fuse physical constraints with temporal patterns, achieving an overall accuracy of 0.9430. The value of the physics-informed approach is further substantiated by its performance in cross-domain generalization, few-shot adaptation, and robustness tests, confirming its potential for practical applications.</div></div>","PeriodicalId":8660,"journal":{"name":"Automation in Construction","volume":"181 ","pages":"Article 106600"},"PeriodicalIF":11.5,"publicationDate":"2025-10-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145314978","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}
Jordyn Golden , Mahsa Mahdavisharif , Gabriel Castelblanco , Giulio Mangano , Andrea Ferrari
{"title":"Disruptions and resilience in construction supply chains","authors":"Jordyn Golden , Mahsa Mahdavisharif , Gabriel Castelblanco , Giulio Mangano , Andrea Ferrari","doi":"10.1016/j.autcon.2025.106577","DOIUrl":"10.1016/j.autcon.2025.106577","url":null,"abstract":"<div><div>Construction Supply Chains (CSCs) frequently face disruptions with relevant effects on inventory backlogs, price instability, and project performance. To address the existing fragmentation on CSC disruptions research, this paper proposes a Systematic Literature Review (SLR) enriched by advanced Natural Language Processing (NLP) techniques to capture how these supply chains have reacted to disruptions and understand how stakeholders can systematically apply emerging technologies and data-driven approaches to optimize CSCs, contributing to advancing intelligent systems and process automation across the construction lifecycle. This review is grounded on foundational resilience frameworks and disruption typologies to contextualize the shift toward integrated technology-enabled CSC and uncover research grouped into macro-level (e.g., transportation-centric) and micro-level (e.g., contractor resilience) clusters. This study advances the field by integrating SLR rigor with scalable NLP to produce a multidimensional framework for CSC resilience, mapping disruption themes and strategic responses across the CSC lifecycle, providing actionable insights for researchers and practitioners.</div></div>","PeriodicalId":8660,"journal":{"name":"Automation in Construction","volume":"181 ","pages":"Article 106577"},"PeriodicalIF":11.5,"publicationDate":"2025-10-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145314982","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}
Changyi Zhao , Hua Bai , Lu Li , Konghua Yang , Zihe Yang , Wankai Chen , Chunbao Liu
{"title":"Dual-arm cooperative construction robots for wall repair with adaptive spraying and troweling","authors":"Changyi Zhao , Hua Bai , Lu Li , Konghua Yang , Zihe Yang , Wankai Chen , Chunbao Liu","doi":"10.1016/j.autcon.2025.106613","DOIUrl":"10.1016/j.autcon.2025.106613","url":null,"abstract":"<div><div>Plastering in construction involves two key processes: spraying and troweling. Most existing robotic systems can only perform one task, leading to low efficiency and inconsistent surface quality. This paper proposes an integrated dual-arm robotic system that unifies spraying, troweling, defect detection, and real-time repair into a continuous and automated workflow. Firstly, an improved detection model, based on YOLOv8 with Hybrid Attention Transformer (HAT) and Dynamic Snake Convolution (DSCov), increases recognition accuracy by 15.1 %. Secondly, a hybrid force–position control strategy is applied to improve plastering uniformity by adapting to varying wall conditions. Thirdly, a sensor fusion method combining inertial measurement units (IMU) and light detection and ranging (LiDAR) is used to correct deviations caused by uneven flooring. Experimental results show that the robot achieves an average deviation of only 0.38 mm from the target plaster thickness of 10 mm, while surface uniformity is improved with the proposed control and correction methods.</div></div>","PeriodicalId":8660,"journal":{"name":"Automation in Construction","volume":"181 ","pages":"Article 106613"},"PeriodicalIF":11.5,"publicationDate":"2025-10-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145314979","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}
Yu Zhang , Long Chen , Qiuchen Lu , Yang Zou , Xiaer Xiahou , Simon Sølvsten , Craig Hancock
{"title":"Semi-automated localised updating for as-built BIM of piping systems using point cloud data","authors":"Yu Zhang , Long Chen , Qiuchen Lu , Yang Zou , Xiaer Xiahou , Simon Sølvsten , Craig Hancock","doi":"10.1016/j.autcon.2025.106609","DOIUrl":"10.1016/j.autcon.2025.106609","url":null,"abstract":"<div><div>As-designed building information models (BIM) often diverge from as-built conditions, limiting their reliability during the operation and maintenance (O&M). Current research focuses on change detection but lacks a systematic workflow for reliable updates, especially for piping systems with frequent changes and complex geometries. The paper addresses how to establish a semi-automated, end-to-end workflow for localised updating as-designed BIM of piping systems from point cloud data. The workflow applies PointNet++ for segmentation, followed by iterative closest point, random sample consensus, and region-growing for geometry extraction. The proposed BIM updating taxonomy and dedicated pre-judgment updating requirements (PUR) and spatial and topological relationships up-dating (STRU) algorithms identify update requirements and automate parametric updates. Validation through case studies demonstrates the workflow's ability to accurately perform localised updates, reducing the manual workload by approximately 70 %. This practical, scalable solution strengthens O&M by maintaining accurate as-built models and inspires future automated BIM updating research.</div></div>","PeriodicalId":8660,"journal":{"name":"Automation in Construction","volume":"181 ","pages":"Article 106609"},"PeriodicalIF":11.5,"publicationDate":"2025-10-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145314980","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 real-time anomaly detection for aerial building machines using adaptive signal decomposition","authors":"Limao Zhang , Jianzhe Jiang , Jiaqi Wang , Zhonghua Xiao , Feilong Fei","doi":"10.1016/j.autcon.2025.106583","DOIUrl":"10.1016/j.autcon.2025.106583","url":null,"abstract":"<div><div>Aerial Building Machine (ABM) is an integrated system for high-rise construction that faces challenges in real-time anomaly detection during operation. This paper proposes an adaptive detection approach combining Variational Mode Decomposition (VMD) and Wavelet Packet Energy Spectrum (WPES) to enhance operational safety. A real case of a 74-floor, 356-m project in Shenzhen demonstrates feasibility and effectiveness. Results show that: (1) the method adaptively detects anomalies in construction and lifting phases, verified by 12 sets of operational data; (2) it exhibits strong robustness, with maximum variation rates of warning indicators of 3.51 % and 2.48 % under Gaussian noise (1 %–5 %), compared to 8.53 % and 8.59 % for traditional methods; and (3) the proposed method demonstrates significant suppression of baseline drift, confining the variation rates of warning indicators below 5% under random-walk disturbances. The contribution lies in using VMD-based Adaptive WPES (VMD–AWPES) for adaptive signal analysis, enabling robust, intelligent anomaly detection in ABM operation. This paper addresses a gap in real-time anomaly detection and promotes intelligent safety management in high-rise construction.</div></div>","PeriodicalId":8660,"journal":{"name":"Automation in Construction","volume":"181 ","pages":"Article 106583"},"PeriodicalIF":11.5,"publicationDate":"2025-10-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145314981","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}