techtranspТехник транспорта: образование и практикаTransport Technician: Education and Practice2687-10252687-1033Federal state budget establishment additional professional education «Educational and instructional center for railway transportation»10.46684/2687-1033.2025.2.203-211SCUPMBtechtransp-772Research ArticleИНФРАСТРУКТУРА ТРАНСПОРТА: СОЗДАНИЕ, ЭКСПЛУАТАЦИЯ И РАЗВИТИЕTRANSPORT INFRASTRUCTURE: ESTABLISHMENT, OPERATION AND DEVELOPMENTПроблематика искусственного интеллекта на Всемирном тоннельном конгрессе 2024 года: обзорProblematics of Artificial Intelligence at the World Tunnelling Congress 2024: Reviewhttps://orcid.org/0000-0001-8635-232XКонюховД. С.KonyuhovD. S.

Дмитрий Сергеевич Конюхов — доктор технических наук, доцент, руководитель отдела научно-технического сопровождения строительства,

123056, Москва, ул. 2-я Брестская, д. 5

Dmitrij S. Konyuhov - Dr. Sci. (Eng.), Associate Professor, Head of the Department of Scientifi c and Technical Support for Construction,

5 2-nd Brestskaya st., Moscow, 123056

gidrotehnik@inbox.ruМосинжпроектРоссияMosinzhproektRussian Federation20251506202562203211Copyright © Конюхов Д.С., 20252025Конюхов Д.С.Konyuhov D.S.Данная работа распространяется под лицензией Creative Commons Attribution 4.0.This work is licensed under a Creative Commons Attribution 4.0 License.https://www.ttspo.ru/jour/article/view/772

На 50-м Всемирном тоннельном конгрессе нашел отражение современный уровень исследований в области подземного строительства, в том числе в сфере искусственного интеллекта на основе цифровизации.

В основу обзора положены опубликованные материалы конгресса. Тематика искусственного интеллекта отражена в отдельном разделе.

Представленную на конгрессе совокупность докладов можно подразделить на следующие несколько направлений: работа тоннелепроходческих машин; изучение деформаций грунтов и сооружений, мониторинг; обучающие методы; цифровые методы проектирования, строительства и управления эксплуатацией.

Предложена интеллектуальная система поддержки принятия решений по обслуживанию конструктивных дефектов тоннелей с использованием графа знаний и глубокого обучения, представлено экспериментальное исследование применения гиперспектральных изображений для оценки прочности бетона на сжатие. Описано прогнозирование качества окружающих пород на основе неполного набора данных из нескольких источников и применения деревьев простейшей байесовской сети. Рассмотрены проектирование, выбор модели и количественная оценка неопределенности при кондиционировании грунта для механизированной щитовой проходки с применением машинного обучения. Предложен энергоэффективный алгоритм управления вентиляцией тоннеля, сочетающий использование динамической нейронной сети и нечеткого управления. Особое внимание уделено оценке производительности тоннелепроходческих машин и выявлению дефектов эксплуатируемых подземных объектов. В целом сфера представленных исследований в области искусственного интеллекта (машинного обучения) применительно к освоению подземного пространства сравнительно узкая и оставляет много аспектов неохваченными.

The 50th World Tunnelling Congress reflects the state of the art in underground construction research, including artificial intelligence based on digitalisation.

The review is based on the published proceedings of the Congress. The subject of artificial intelligence is reflected in a separate section.

The set of reports presented at the Congress can be divided into the following several areas: operation of tunnel boring machines; study of ground and structure deformations, monitoring; training methods; digital methods of design, construction and operation management.

An intelligent decision support system for the maintenance of structural defects in tunnels using knowledge graph and deep learning is proposed, and an experimental study of the application of hyperspectral images to assess the compressive strength of concrete is presented. Prediction of surrounding rock quality based on an incomplete data set from multiple sources and the application of simple Bayesian network trees is described. The design, model selection and uncertainty quantification of soil conditioning for mechanised tunnel sinking using machine learning are discussed. An energy-efficient tunnel ventilation control algorithm combining the application of dynamic neural network and fuzzy control is recommended. Special attention is paid to the performance evaluation of tunnel boring machines and defect detection of operating underground facilities. In general, the scope of the presented research in the fi eld of artificial intelligence (machine learning) as applied to the development of underground space is relatively narrow and leaves many aspects uncovered.

дефекты эксплуатируемых тоннелейискусственный интеллекткомпьютерное обучениеподземное пространствотоннелепроходческая машинатранспортные тоннелихарактеристики грунтовэффективность строительства тоннелейdefects of operated tunnelsartificial intelligencecomputer learningunderground spacetunnel boring machinetransport tunnelssoil characteristicstunnel construction efficiencyReferencesTunnelling for a Better Life. Taylor & Francis, 2024. Pp. 2995–3299.Tunnelling for a Better Life. Taylor & Francis, 2024;2995-3299.Iasiello C., Rodríguez-Sánchez J. TBM machine parameters estimation: From design approach to on-fi eld results. A concrete example based on Kalman Filter approach // Tunnelling for a Better Life. 2024. Pp. 3023–3028.Iasiello C., Rodríguez-Sánchez J. TBM machine parameters estimation: From design approach to on-fi eld results. A concrete example based on Kalman Filter approach. Tunnelling for a Better Life. 2024;3023-3028.Kim D., Shin Y., Kim D., Lee C., Kwon K. et al. A novel machine-learning model for estimating disc cutter life in TBMs considering individual cutter travel lengths // Tunnelling for a Better Life. 2024. Pp. 3053–3058.Kim D., Shin Y., Kim D., Lee C., Kwon K. et al. A novel machine-learning model for estimating disc cutter life in TBMs considering individual cutter travel lengths. Tunnelling for a Better Life. 2024;3053-3058.Kang Y.S., Park S.J., Hwang J.H., Hong J.P., Ko T.Y. Prediction of disc cutter wear considering ground conditions and TBM operating parameters // Tunnelling for a Better Life. 2024. Pp. 3037–3043.Kang Y.S., Park S.J., Hwang J.H., Hong J.P., Ko T.Y. Prediction of disc cutter wear considering ground conditions and TBM operating parameters. Tunnelling for a Better Life. 2024;3037-3043.CatBoost in Machine Learning // GeeksforGeeks. 2025. URL: https://www.geeksforgeeks.org/catboost-mlCatBoost in Machine Learning. GeeksforGeeks. 2025. URL: https://www.geeksforgeeks.org/catboost-mlMiller M., Fang Y., Luo H., Wang Y., Xu G. et al. Forecasting the driving speed of the TBM using machine learning Algorithms // Tunnelling for a Better Life. 2024. Pp. 3067–3072.Miller M., Fang Y., Luo H., Wang Y., Xu G. et al. Forecasting the driving speed of the TBM using machine learning Algorithms. Tunnelling for a Better Life. 2024;3067-3072.Dardashti A., Rostami J., Ajalloeian R., Hassanpour J., Salimi A. Development of a hard rock TBM performance prediction model using RMR input parameters // Tunnelling for a Better Life. 2024. Pp. 2995–3004.Dardashti A., Rostami J., Ajalloeian R., Hassanpour J., Salimi A. Development of a hard rock TBM performance prediction model using RMR input parameters. Tunnelling for a Better Life. 2024;2995-3004.Dewangan A., Sahoo D.R., Karlovsek J. Performance analysis of supervised algorithms on encoded data for predicting tunnel strain classes // Tunnelling for a Better Life. 2024. Pp. 3005–3013.Dewangan A., Sahoo D.R., Karlovsek J. Performance analysis of supervised algorithms on encoded data for predicting tunnel strain classes. Tunnelling for a Better Life. 2024;3005-3013.Feng Y., Zhang X., Feng S., Zhao Y., Chen Y. Automatic classification and segmentation of tunnel cracks based on deep learning and visual explanations // Tunnelling for a Better Life. 2024. Pp. 3014–3022.Feng Y., Zhang X., Feng S., Zhao Y., Chen Y. Automatic classifi cation and segmentation of tunnel cracks based on deep learning and visual explanations. Tunnelling for a Better Life. 2024;3014-3022.Wang Y., Yang L., Liu X., Yan P. An improved semantic segmentation algorithm for high-resolution remote sensing images based on DeepLabv3+ // Scientific Reports. 2024. Vol. 14. Issue 1. DOI: 10.1038/s41598-024-60375-1Wang Y., Yang L., Liu X., Yan P. An improved semantic segmentation algorithm for high-resolution remote sensing images based on DeepLabv3+. Scientific Reports. 2024;14(1). DOI: 10.1038/s41598-024-60375-1Jia F., Xue Y., Zhang Q., Qu L. An intelligent decision support system for tunnel structural defects maintenance with combining knowledge graph and deep learning // Tunnelling for a Better Life. 2024. Pp. 3029–3036.Jia F., Xue Y., Zhang Q., Qu L. An intelligent decision support system for tunnel structural defects maintenance with combining knowledge graph and deep learning. Tunnelling for a Better Life. 2024;3029-3036.Ultralytics YOLOv8 // Хабр. 2023. URL: https://habr.com/ru/articles/710016Ultralytics YOLOv8. Habr. 2023. URL: https://habr.com/ru/articles/710016 (In Russ.).Katuwal T.B., Panthi K.K., Basnet C.B., Adhikari S. Leakage prediction and post-grouting assessment in headrace tunnel of a hydropower project // Tunnelling for a Better Life. 2024. Pp. 3044–3052.Katuwal T.B., Panthi K.K., Basnet C.B., Adhikari S. Leakage prediction and post-grouting assessment in headrace tunnel of a hydropower project. Tunnelling for a Better Life. 2024; 3044-3052.Murro V.D., Ouyang A., Osborne J.A., Li Z. Intelligent tunnel asset management of CERN underground facilities // Tunnelling for a Better Life. 2024. Pp. 3073–3078.Murro V.D., Ouyang A., Osborne J.A., Li Z. Intelligent tunnel asset management of CERN underground facilities. Tunnelling for a Better Life. 2024;3073-3078.Wang C., Huang H., Zhou M., Zhu S. Hyperspectral imaging features for concrete compressive strength assessment: Experimental study // Tunnelling for a Better Life. 2024. Pp. 3104–3112.Wang C., Huang H., Zhou M., Zhu S. Hyperspectral imaging features for concrete compressive strength assessment: Experimental study. Tunnelling for a Better Life. 2024;3104-3112.Никонов А.В., Давлетшин Р.В., Яковлева Н.И., Лазарев П.С. Фильтрация методом Савицкого – Голея спектральных характеристик чувствительности матричных фотоприемных устройств // Успехи прикладной физики. 2016. Т. 4. № 2. С. 198–205. EDN VXBTOL.Nikonov A.V., Davletshin R.V., Iakovleva N.I., Lazarev P.S. Savitzky-Golay smoothing method of FPA photodiodes spectral response. Advances in Applied Physics. 2016;4(2):198-205. EDN VXBTOL. (In Russ.).Wu C., Huang H., Tong H., Zhou M., Zhang L. et al. Investigation on surrounding rock quality prediction based on incomplete multi-source dataset and tree-augmented naive Bayesian network // Tunnelling for a Better Life. 2024. Pp. 3131–3138.Wu C., Huang H., Tong H., Zhou M., Zhang L. et al. Investigation on surrounding rock quality prediction based on incomplete multi-source dataset and tree-augmented naive Bayesian network. Tunnelling for a Better Life. 2024;3131-3138.Обзор архитектуры Swin Transformer // Хабр. 2022. URL: https://habr.com/ru/articles/599057Overview of Swin Transformer architecture. Habr. 2022. URL: https://habr.com/ru/articles/599057 (In Russ.).Liu Z., Lin Y., Cao Y., Hu H., Wei Y. et al. Swin Transformer: Hierarchical Vision Transformer using Shifted Windows // 2021 IEEE/CVF International Conference on Computer Vision (ICCV). 2021. Pp. 9992–10002. DOI: 10.1109/ICCV48922.2021.00986Liu Z., Lin Y., Cao Y., Hu H., Wei Y. et al. Swin Transformer: Hierarchical Vision Transformer using Shifted Windows. 2021 IEEE/CVF International Conference on Computer Vision (ICCV). 2021;9992-10002. DOI: 10.1109/ICCV48922.2021.00986Wang M., Zhao S., Yi W., Peng X. Intelligent surrounding rock classification and mechanical parameters analysis method based on drilling parameters of tunnels // Tunnelling for a Better Life. 2024. Pp. 3122–3130.Wang M., Zhao S., Yi W., Peng X. Intelligent surrounding rock classification and mechanical parameters analysis method based on drilling parameters of tunnels. Tunnelling for a Better Life. 2024;3122-3130.Lin W., Sheil B., Xie X., Li K., Niu G. Segment segmentation of tunnel ring point clouds using 3D deep learning // Tunnelling for a Better Life. 2024. Pp. 3059–3066.Lin W., Sheil B., Xie X., Li K., Niu G. Segment segmentation of tunnel ring point clouds using 3D deep learning. Tunnelling for a Better Life. 2024;3059-3066.Qi C.R., Yi L., Su H., Guibas L.J. PointNet++: Deep Hierarchical Feature Learning on Point Sets in a Metric Space. 2017. DOI: 10.48550/arXiv.1706.02413Qi C.R., Yi L., Su H., Guibas L.J. PointNet++: Deep Hierarchical Feature Learning on Point Sets in a Metric Space. 2017. DOI: 10.48550/arXiv.1706.02413Samadi H., Hassanpour J., Rostami J., Moghbeli A. Assessment of TBM performance in different types of rocks using supervised learning techniques // Tunnelling for a Better Life. 2024. Pp. 3095–3103.Samadi H., Hassanpour J., Rostami J., Moghbeli A. Assessment of TBM performance in different types of rocks using supervised learning techniques. Tunnelling for a Better Life. 2024;3095-3103.Yuan X., Wang S., Qu T. Machine learning-informed soil conditioning for mechanized shield tunneling feature engineering, model selection, and uncertainty quantification // Tunnelling for a Better Life. 2024. Pp. 3139–3145.Yuan X., Wang S., Qu T. Machine learning-informed soil conditioning for mechanized shield tunneling feature engineering, model selection, and uncertainty quantification. Tunnelling for a Better Life. 2024;3139-3145.Pei L., Wu H., Hu M., Lu J., Wu B. et al. Research and practice of digital lean construction mode of tunnelling based on shield self-driving technology // Tunnelling for a Better Life. 2024. Pp. 3079–3085.Pei L., Wu H., Hu M., Lu J., Wu B. et al. Research and practice of digital lean construction mode of tunnelling based on shield self-driving technology. Tunnelling for a Better Life. 2024;3079- 3085.Sakai K., Miyanaga S., Yamagami M. Study on machine learning method for supporting conventional tunnel engineering judgement // Tunnelling for a Better Life. 2024. Pp. 3086–3094.Sakai K., Miyanaga S., Yamagami M. Study on machine learning method for supporting conventional tunnel engineering judgement. Tunnelling for a Better Life. 2024;3086-3094.Wang H., Li Z., Zhang Y., Zhang J. An energy-effi cient tunnel ventilation control algorithm combining dynamic neural network and fuzzy control // Tunnelling for a Better Life. 2024. Pp. 3113–3121.Wang H., Li Z., Zhang Y., Zhang J. An energy-effi cient tunnel ventilation control algorithm combining dynamic neural network and fuzzy control. Tunnelling for a Better Life. 2024:3113-3121.

The authors declare that there are no conflicts of interest present.