Intelligent predictive quality control systems in construction: analyzing the effectiveness of machine learning for automated inspection of construction defects
How to cite
Abstract
The modern construction industry is characterized by a high level of defects, reaching 15-30% of the total volume of work, due to the limitations of traditional methods of visual quality control. The present study presents a comprehensive analysis of the use of intelligent predictive quality control systems based on machine learning algorithms for automated detection of construction defects. The aim of the work is to evaluate the effectiveness of various neural network architectures in the tasks of classifying and segmenting defects in building structures with the development of methodological recommendations for their practical implementation. The study was conducted using convolutional neural networks ResNet-50, U-Net and YOLOv5 on a dataset of 12,847 images of construction defects collected from 73 objects for various purposes in the period 2023-2024. Methods of cross-validation, statistical analysis and economic efficiency assessment were used. The results demonstrate that the modified U-Net architecture provides 89.4% crack detection accuracy in concrete structures compared to 62.1% with traditional inspection methods, while reducing the inspection time from 4.2 hours to 0.8 hours per 1000 m2 of area. The economic analysis showed a 67.3% reduction in quality control costs while increasing the detection of critical defects by 73.6%. The practical significance of the work lies in the formation of scientifically based recommendations on the selection of optimal machine learning architectures for various types of construction sites, which contributes to improving the quality and safety of construction while reducing economic costs.
Keywords
References
Bezerra A., da Silva K.C.N., Nascimento E. Industrial environment: A strategy for preventive maintenance using machine learning to predict the useful life of equipment and statistical process control for continuous monitoring of variables // INFOCOMP journal of computer science. 2023. Vol. 22. pp. 1-9.
Bianchi E., Hebdon M. Visual structural inspection datasets // Automation in Construction. 2022. Vol. 139. pp. 104-299.
Cardellicchio A., Nobile S., Tateo A., Zaccara A., Lomonte G., Angelastro A. Physical interpretation of machine learning-based recognition of defects for the reliability assessment of welded joints // Electronic markets. 2024.
Dais D., Bal I.E., Smyrou E., Sarhosis V. Automatic crack classification and segmentation on masonry surfaces using convolutional neural networks and transfer learning // Automation in construction. 2021. Vol. 125. pp. 103-606.
Dorafshan S., Thomas R.J., Maguire M. SDNET2018: An annotated image dataset for non-contact concrete crack detection using deep convolutional neural networks // Data in brief. 2018. Vol. 21. pp. 1664-1668.
Elhariri E., El-Bendary N., Taie S.A. Historical-crack18-19: A dataset of annotated images for non-invasive surface crack detection in historical buildings // Data in brief. 2022. Vol. 41. pp. 107-865.
Guo J., Liu P., Xiao B., Deng L., Wang Q. Surface defect detection of civil structures using images: Review from data perspective // Automation in сonstruction. 2024. Vol. 158. рр. 105-186.
He Z., Chen J., Zhang Y.-H. Wang Infrastructure Crack Segmentation: Boundary Guidance Method and Benchmark Dataset // arXiv preprint arXiv:2306.09196. 2023.
Yang G., Liu K., Zhang J., Zhao B., Zhao Z., Chen X., Chen B.M. Datasets and processing methods for boosting visual inspection of civil infrastructure: A comprehensive review and algorithm comparison for crack classification, segmentation, and detection // Construction and building materials. 2022. Vol. 356. pp. 129-226.
Yang L., Li B., Li W., Liu Z., Yang G., Xiao J. A robotic system towards concrete structure spalling and crack database // 2017 IEEE Inter. conf. on robotics and biomimetics (ROBIO). 2017. pp. 1276-1281.
Yokoyama S., Matsumoto T. Development of an automatic detector of cracks in concrete using machine learning // Procedia engineering. 2017. Vol. 171. pp. 1250-1255.
Zhang G., Pan Y., Zhang L. Semi-supervised learning with GAN for automatic defect detection from images // Automation in Construction. 2021. Vol. 128. pp. 103-764.