Oct 2026· Journal of construction engineering and management· Vol 152· 0 citations· 53 references
Abstract
This paper proposes a vision-based framework named concrete progress and concrete health monitoring (CPCHM) for automated compliance supervision and health risk assessment during concrete pouring and vibration operations. The framework integrates YOLOv8-pose for worker posture estimation and YOLOv8-detection for equipment identification, achieving average accuracies of 92.8% and 99.5%, respectively. By fusing posture and equipment features, a support vector machine classifier distinguishes between pouring and vibration operations with 90.8% accuracy and an F1-score of 90.3%. Further, a spatiotemporal graph convolutional network is employed to model elbow joint dynamics and assess musculoskeletal risks, reaching a behavioral classification accuracy of 89.6%. To address occlusion and multiworker collaboration, CPCHM introduces a distance-based operator identification method and an adaptive region-of-interest inference mechanism, maintaining stable keypoint tracking and continuous elbow-angle estimation even under partial visibility. The framework is embedded on a Jetson nano edge device, which automatically triggers an acoustic buzzer alert when pouring durations exceed 90 min or vibration times fall outside the 5–15 s standard. CPCHM provides a compact, sensor-free, and scalable solution for integrating progress compliance monitoring and ergonomic health assessment, enabling intelligent and real-time supervision in dynamic concrete construction environments.
Structural health monitoring systems based on machine learning routinely achieve high classification accuracy but rarely explain the basis of their decisions, limiting their adoption by practicing engineers who must justify safety-critical actions. This paper presents a framework that combines vibration-based and image-based damage assessment with explainable artificial intelligence and a data-driven digital twin. An XGBoost classifier is trained on physics-informed modal features extracted from 15 months of monitoring data from the KW51 railway bridge in Leuven, Belgium, to distinguish healthy, damaged, and repaired structural states, while Shapley Additive Explanations are used to attribute each prediction to specific frequency, damping, and environmental features. In parallel, an EfficientNet-B0 convolutional neural network is fine-tuned on the Concrete Defect Bridge Image dataset to classify six concrete surface defect types, and Gradient-weighted Class Activation Mapping is used to visually localize the image regions driving each defect prediction. The outputs of the two branches are combined into a Structural Health Index designed to support a continuously updated data-driven digital twin. On a temporally held-out test set, the vibration classifier achieves a macro F1 score of 0.682 and an area under the receiver operating characteristic curve of 0.880, while the image classifier achieves a macro F1 score of 0.886 and an area under the curve of 0.977. The study also identifies three findings with direct relevance for deployment: environmental variables retain a measurable share of feature importance after detrending, the vibration classifier confuses the damaged and repaired classes in a safety-relevant direction, and a fixed-weight fusion of the two branches on unpaired data does not improve upon the vibration branch alone. Together, these findings highlight the importance of paired multi-modal datasets for reliable fusion and provide practical guidance for the development of robust, explainable, and validated digital twin frameworks for structural health monitoring.
Unknown authors· Frontiers in Built Environme...· 0 citations
An intelligent safety belt model with self-identification and self-sensing capabilities is developed, which significantly improves safety supervision for work at height and provides proactive protection for personnel performing elevated tasks.
Wending Li, Jian-Lun Lin, Minghui Lin et al.· International Conference on...· 0 citations
The proposed Media Pipe–CNN framework provides an efficient, accurate, and marker less solution for automated ergonomic risk assessment, supporting intelligent occupational safety management, continuous workplace monitoring, and the implementation of smart manufacturing systems aligned with Industry 4.0 initiatives.
Rahmadwati Rahmadwati, Farrel Rafif Ferdian, Y. Sumantri et al.· International Journal of Eng...· 0 citations
Computer-vision-based structural health monitoring (CVSHM) enables contactless displacement measurement at multiple locations on the vibrating structure. Additionally, such a measurement can be realized from a certain distance from the monitored infrastructure. It provides a possibility of reducing the costs of the CVSHM system. However, computer-vision-based (CV) vibration measurement can be significantly contaminated by measurement errors at higher vibration frequencies and low amplitudes.
A framework for CVSHM is proposed, which allows for robust detection, localization, and assessment of the damage [1]. As illustrated in the figure, the framework consists of: CV measurement of structural displacements, modal data extraction, and modal sensitivity-based model updating by changing the stiffness of monitored structural members. Displacement measurement is realized with the template matching technique, maximizing the zero-normalized cross-correlation function. Later, modal parameters are identified using the data-driven stochastic subspace identification method (SSI-DATA). Degrees of freedom suspected to be source of gross errors are removed. The last framework component is inspired by the augmented inverse estimate described in [2]. The proposed method additionally employs: (a) weighting of errors between the identified and model modal parameters, and (b) constraints imposed on the model updating procedure that structural stiffness parameters can only decrease with respect to the initial values. Such a method of damage assessment is termed weighted negative least square inverse estimate (WNLSIE). This method is physics-based and avoids the problems typical of machine-learning approaches, such as the need for the collection of training data or generalization of the trained model.
The proposed framework is tested using realistic synthetic videos representing vibrating truss structure. These videos are generated with physics-based graphical models (PBGM) and allow for employing the displacement ground truth data for comparison purposes [3]. Displacements of 19 truss nodes are measured and 29 truss members are monitored. Sampling frequency is 120 frames per second (fps). Excitations are unknown. The proposed framework allows for prediction of damaged truss member and its damage level, when error of measured displacement is at the level of 50 %.
The proposed framework for CVSHM is easy to implement and allows for detection, localization, and assessment of structural damage even for highly contaminated displacement data.
Bibliography
[1] M. Ostrowski, B. Blachowski, B. Wójcik, M. Żarski, P. Tauzowski, Ł. Jankowski, A framework for computer vision-based health monitoring of a truss structure subjected to unknown excitations, Earthquake Engineering and Engineering Vibration 22 (2023) 1–17. https://doi.org/10.1007/s11803-023-2154-3.
[2] X. Peng, F.J. Qin, Q.W. Yang, H. Chen, A Robust Estimate Method for Damage Detection of Concrete Structures Using Contaminated Data, Advances in Civil Engineering 2021 (2021) 6669958. https://doi.org/https://doi.org/10.1155/2021/6669958.
[3] Y. Narazaki, F. Gomez, V. Hoskere, M.D. Smith, B.F. Spencer, Efficient development of vision-based dense three-dimensional displacement measurement algorithms using physics-based graphics models, Structural Health Monitoring 20 (2021) 1841–1863. https://doi.org/https://doi.org/10.1177/1475921720939522.
M. Ostrowski, B. Błachowski, M. Żarski et al.· e-Journal of Nondestructive...· 0 citations
Workplaces in construction and industry suffer from a significant number of workplace accidents because of a lack of safety mechanisms like helmets and high-visibility vests. In order to solve this issue, an automatic detection system for detecting the presence of the worker's helmet & safety vest using a deep learning model is developed. You Only Look Once – Neural Architecture Search (YOLO-NAS) algorithm was chosen to be used in the model because of its fast and high-quality detection process. A dataset with images containing healthy workers wearing helmets & safety vests (with some workers appearing in different safety vests) is collected and used as the training set. The resulting model is then saved for future usage in making predictions with new images. Then, this model is connected to Streamlit, which provides a convenient Web-based user interface through which a user can provide an image as an input to the model. The trained neural network is applied to the input image, and all detected objects that correspond to helmet & safety vest are put in bounding boxes on the input image. Moreover, each of the bounding boxes is labeled with the name of the object and confidence score of the detection. Thus, a clear visualization of the safety equipment worn by the worker is provided. Overall, the developed detection system significantly decreases the human intervention needed for visual verification of the safety helmet or vest. The described system is a user-friendly and cost-effective way of evaluating workplace safety compliance.
S. Vijayakumar, Loganathan Nachimuthu, Balasubramaniam C et al.· 2026 International Conferenc...· 0 citations
Accurate road crack detection is essential for intelligent pavement inspection, yet thin crack morphology, cluttered backgrounds, and deployment constraints still challenge lightweight detectors. This paper presents an improved YOLOv11s-based detector for road distress recognition. Three coordinated modules are introduced: a C3k2- SHSA-CGLU backbone block for stronger contextual perception and dynamic crack-feature filtering, a GLSABiFPN neck for bidirectional multi-scale fusion with enhanced fine-detail retention, and a lightweight shared-convolution detection head for compact prediction. Experiments on the China subset of RDD2022 show that the proposed method improves mAP@0.5 from 87.2% to 89.4% and reduces parameters from 9.41 M to 7.32 M compared with YOLOv11s. Additional cross-dataset results on GRDDC2020 indicate acceptable generalization, while the reduced parameter count and compact model size suggest good deployment potential. Overall, the method provides a practical balance between detection accuracy and model compactness for automated pavement inspection.
Shaowen Zhang, Mengjuan Chen, Liejun Wang et al.· International Conference on...· 0 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.
