Bridging the Simulation-to-Field Gap in AI-Based Underground Mine Inspection: A Systematic Review of Synthetic Data, Domain Adaptation, Transfer Learning, and Digital Twins
Article Sidebar
Main Article Content
Abstract
Artificial intelligence (AI) has advanced rapidly in underground and tunnel inspection, but field deployment continues to be limited by a persistent simulation-to-field reality gap. Real underground data are costly to acquire, difficult to annotate, safety-sensitive, highly site-specific, and often scarce for rare but consequential defects. Simulation, synthetic data, transfer learning, domain adaptation, and digital twins offer complementary ways to address this constraint, yet these research streams are usually reviewed in isolation. This systematic review treats field generalisation as the central analytical problem and synthesises the full pathway from virtual environment construction and synthetic-data generation through cross-domain adaptation, limited-data fine-tuning, digital-twin calibration, and field validation. We used a PRISMA-informed protocol to review literature published from 2015 to 20 August 2026, classifying evidence into three levels: direct underground-mine studies, neardomain tunnel/subterranean studies, and transferable sim-to-real methodology. The synthesis indicates that the most convincing recent results arise not from simulation fidelity alone, but from combining synthetic diversity with target-domain adaptation. Geometry-informed tunnel point-cloud simulation, for example, has shown strong synthetic-only performance and further gains after limited-real-data transfer learning. At the same time, domain-adaptive crack-segmentation studies report substantial performance recovery across materials, structures, and field sites. Mining digital twins are also moving beyond static visualisation toward calibrated virtual-real loops, including mine-tunnel reconstruction, ventilation modelling, rescue-robot navigation, and field-tested teleoperation. Nevertheless, cross-mine validation, explicit measurement of the reality gap, rare-hazard simulation, uncertainty-aware adaptation, and standardised deploymentreadiness criteria remain weak. To address these gaps, this review proposes an Underground Simulation-to-Field Gap Taxonomy and the MineSim2Field framework, an eight-stage closed-loop pathway connecting virtual environment construction, synthetic data, domain diversification, pretraining, adaptation, field calibration, deployment, and twin recalibration. The framework positions simulation as a continuously updated source of evidence rather than a one-off substitute for field data.
Downloads
Article Details
Section
This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.
How to Cite
References
Zhang, W., Tian, C., Yang, J., Xiao, P., Wu, Z., Li, H., & Sun, W. (2026). Vision and deep learning-based tunnel surface defect detection: Technological evolution, challenges, and prospects. Measurement, 257, 118853. DOI: 10.1016/j.measurement.2025.118853
Xue, Y.-D., Jia, F., Luo, W., Zhang, D.-M., Liu, J., & Guo, Y.-F. (2026). Deep learning-based computer vision methods for shield tunnel defect recognition—automation in Construction, 183, 106807. DOI: 10.1016/j.autcon.2026.106807
Nobahar, P., Xu, C., Dowd, P., & Shirani Faradonbeh, R. (2024). Exploring digital twin systems in mining operations: A review. Green and Smart Mining Engineering, 1(4), 474–492. DOI: 10.1016/j.gsme.2024.09.003
Ebad, S. A., Abueid, A. I., Amara, M., & Ahmed, R. (2026). AI-Driven Digital Twins in Mining Operations: A Comprehensive Review. Technologies, 14(5), 269. DOI: 10.3390/technologies14050269
Liang, R., Zhang, C., Li, B., Saydam, S., & Canbulat, I. (2026). Evolution of visualisation and digital twin technologies in mining. International Journal of Coal Science & Technology, 13, 53. DOI: 10.1007/s40789-026-00866-w
Muratore, F., Ramos, F., Turk, G., Yu, W., Gienger, M., & Peters, J. (2022). Robot Learning From Randomised Simulations: A Review. Frontiers in Robotics and AI, 9, 799893. DOI: 10.3389/frobt.2022.799893
Tobin, J., Fong, R., Ray, A., Schneider, J., Zaremba, W., & Abbeel, P. (2017). Domain randomisation for transferring deep neural networks from simulation to the real world. 2017 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), 23–30.
DOI: 10.1109/IROS.2017.8202133
Ganin, Y., Ustinova, E., Ajakan, H., Germain, P., Larochelle, H., Laviolette, F., Marchand, M., & Lempitsky, V. (2016). Domain-adversarial training of neural networks. Journal of Machine Learning Research, 17(59), 1–35.
Sadeghi, F., & Levine, S. (2017). CAD2RL: Real single-image flight without a single real image. Robotics: Science and Systems XIII.
DOI: 10.15607/RSS.2017.XIII.034
Tiboni, G., Arndt, K., & Kyrki, V. (2023). DROPO: Sim-to-real transfer with offline domain randomisation. Robotics and Autonomous Systems, 166, 104432. DOI: 10.1016/j.robot.2023.104432
Pan, S. J., & Yang, Q. (2010). A survey on transfer learning. IEEE Transactions on Knowledge and Data Engineering, 22(10), 1345–1359. DOI: 10.1109 /TKDE.2009.191
Weng, X., Huang, Y., Li, Y., Yang, H., & Yu, S. (2023). Unsupervised domain adaptation for crack detection. Automation in Construction, 153, 104939. DOI: 10.1016/j.autcon.2023.104939
Shim, S. (2025). Self-supervised domain adaptive approach for extrapolated crack segmentation with fine-tuned inpainting generative model. Computer-Aided Civil and Infrastructure Engineering, 40, 3333–3358. DOI: 10.1111/mice.13517
Yang, W., Lin, W., Jing, Y., Hill, T., Wysocki, O., & Sheil, B. (2026). Tunnel scanner: Geometry-informed synthetic point cloud generation and transfer learning for tunnel segmentation. Automation in Construction, 187, 106924. DOI: 10.1016/j.autcon.2026.106924
Zhao, Y., Jiao, S., Yang, T., Wang, Y., Xu, H., & Liu, Z. (2026). A novel transfer learning-enhanced BiLSTM-DCNN architecture for mine microseismic signal identification with a small dataset. Geomechanics and Geophysics for Geo-Energy and Geo-Resources, 12, 104. DOI: 10.1007/s40948-026-01168-w
Jacobs, D. R., van Laar, J. H., & Schutte, C. S. L. (2022). Strategy to identify and mitigate hazards in deep-level mine ventilation systems using a calibrated digital twin. South African Journal of Industrial Engineering, 33(3), 175–191. DOI: 10.7166/33-3-2795
Zhu, H., Jin, J., & Zhao, S. (2025). Research on Multi-Sensor Data Fusion Based Real-Scene 3D Reconstruction and Digital Twin Visualization Methodology for Coal Mine Tunnels. Sensors, 25(19), 6153. DOI: 10.3390/s25196153
Cui, Y., Li, J., Li, Y., Zhang, Z., & Du, J. (2026). Multiscale geometric digital twins of mine tunnels that integrate topological evolution and local reconstruction. Tunnelling and Underground Space Technology, 169, 107317. DOI: 10.1016/j.tust.2025.107317
You, S., Li, M., Wu, B., Wang, J., & Tang, C. (2026). Digital Twin of Coal Mine Rescue Robot—Research on Intelligence and Visualisation. Sensors, 26(9), 2840. DOI: 10.3390/s26092840
Page, M. J., McKenzie, J. E., Bossuyt, P. M., Boutron, I., Hoffmann, T. C., Mulrow, C. D., et al. (2021). The PRISMA 2020 statement: An updated guideline for reporting systematic reviews. BMJ, 372, n71. DOI: 10.1136/bmj.n71
Zhuge, H., Lee, X. Y., Sin, G., Ahmed, R., Vidyaratne, L., Kumar, A., & Farahat, A. (2026). Active Sim-to-Real Gap Reduction for Industrial Inspection via Digital Twin and Embedding Analysis. PHM Society European Conference, 9(1), 1–11.
DOI: 10.36001/phme.2026.v9i1.5037
Zhou, Z., Zhang, J., Gong, C., & Wu, W. (2023). Automatic tunnel lining crack detection via deep learning with generative adversarial network-based data augmentation. Underground Space, 9, 140–154. DOI: 10.1016/j.undsp.2022.07.003
Shim, S. (2024). Self-training approach for crack detection using synthesised crack images based on a conditional generative adversarial network. Computer-Aided Civil and Infrastructure Engineering, 39, 1019–1041. DOI: 10.1111/mice.13119
Yan, Q., Qiao, M., Liao, X., & Xiong, Z. (2026). Intelligent tunnel lining defect identification via a parallel multiscale attention convolutional neural network with GAN-assisted data augmentation. Structural Concrete. DOI10.1002 /sucoco.70489
Han, W., Xiao, Y., & Yin, Y. (2024). UM-GAN: Underground mine GAN for underground mine low-light image enhancement. IET Image Processing, 18(8), 2154–2160. DOI: 10.1049/ipr2.13092
Xie, J., Chen, B., Giacomini, A., Guo, H., Iqbal, U., & Huang, J. (2025). A versatile synthetic data generation framework for crack detection: Engineering Structures, 344, 121428. DOI: 10.1016/j.engstruct.2025.121428
Liao, X., Yan, Q., Xiang, C., Qiao, M., Li, B., & Zhang, C. (2025). Crack severity assessment of shield tunnel segments based on DFOS strain measurements assisted by deep learning. Measurement, 256, 118332. DOI: 10.1016/j.measurement.2025.118332
Kritzinger, W., Karner, M., Traar, G., Henjes, J., & Sihn, W. (2018). Digital Twin in manufacturing: A categorical literature review and classification. IFAC-PapersOnLine, 51(11), 1016–1022. DOI: 10.1016/j.ifacol.2018.08.474
Machado, L. B., Poncetti, B. L., de Assis, L. S., & Futai, M. M. (2025). Automated procedural modelling for tunnel inspection and digital twin development. Tunnelling and Underground Space Technology, 165, 106929. DOI: 10.1016/j.tust.2025.106929
Sun, H., Zhu, M., Dai, Y., Li, X., Rui, Y., Zhu, W., & Zhu, H. (2025). Digital twin of advanced geological conditions in rock tunnel construction—automation in Construction, 178, 106407. DOI: 10.1016/j.autcon.2025.106407
Cacciuttolo, C., Atencio, E., Komarizadehasl, S., & Lozano-Galant, J. A. (2025). Development of an Advanced Multi-Layer Digital Twin Conceptual Framework for Underground Mining. Sensors, 25(21), 6650. DOI: 10.3390/s25216650
Cao, P., Liu, J., Wang, H., Wang, Y., Liu, X., et al. (2024). Uncertainty analysis of a digital twin model of a mine ventilation system. Scientific Reports, 14, 26558. DOI: 10.1038/s41598-024-77978-3
Aksoy, C. O., Uyar Aksoy, G. G., Aydemir, Y., Sarısan, H. B., & Kaya, E. (2025). Use of Digital Twin Application Performed with CFD Analysis in an Underground Mine to Interpret Events During and After a Mine Fire. Mining, 5(4), 75. DOI: 10.3390/mining5040075
Le, D. T., Sutjipto, S., Nguyen, D. D. K., & Paul, G. (2025). Design, Integration, and Field Testing of a Digital Twin-Based Teleoperated Rock Scaling Robot. IEEE Transactions on Field Robotics. DOI: 10.1109/TFR.2025.3547807
Hoffman, J., Tzeng, E., Park, T., Zhu, J.-Y., Isola, P., Saenko, K., Efros, A., & Darrell, T. (2018). CyCADA: Cycle-consistent adversarial domain adaptation. Proceedings of the 35th International Conference on Machine Learning, 80, 1989–1998.
Beyene, D. A., Tran, D. Q., Maru, M. B., Kim, T., Park, S., & Park, S. (2023). Unsupervised domain adaptation-based crack segmentation using a transformer network. Journal of Building Engineering, 79, 107889. DOI: 10.1016/j.jobe.2023.107889
Bhattacharya, S., Zhang, C., Mishra, D. K., Yuen, M. M. F., & Zhang, J. (2025). Efficient unsupervised domain adaptation for crack segmentation with interpretable Fourier–Morphology blending and uncertainty-guided self-training. Computer-Aided Civil and Infrastructure Engineering, 40, 5790–5807. DOI: 10.1111/mice.70127
Maharjan, S., Inadomi, S., Itakura, K., & Chun, P.-J. (2025). Domain-adaptive self-supervised learning for corrosion detection and 3D building information model mapping in steel tunnels. Computer-Aided Civil and Infrastructure Engineering, 40, 4425–4447.
DOI: 10.1111/mice.70077
Liu, K., & Chen, B. M. (2023). Industrial UAV-Based Unsupervised Domain Adaptive Crack Recognition: From Database Towards Real-Site Infrastructural Inspections. IEEE Transactions on Industrial Electronics, 70(9), 9410–9420. DOI: 10.1109/TIE.2022.3204953
Tanveer, M., & Cho, S. (2026). Unsupervised domain adaptation framework for crack segmentation using consistency regularisation based on cross-domain image augmentation. Computer-Aided Civil and Infrastructure Engineering, 42, 100036.
DOI: 10.1016/j.cacaie.2026.100036
Liu, F., Ding, W., Qiao, Y., & Wang, L. (2024). Transfer learning-based encoder-decoder model with visual explanations for infrastructure crack segmentation: New open database and comprehensive evaluation. Underground Space, 17, 60–81.
DOI: 10.1016/j.undsp.2023.09.012
Bao, X., Huang, J., Shen, J., Wu, X., Wang, T., Chen, X., & Cui, H. (2025). A novel method of void detection in rebar-affected areas based on transfer learning and improved YOLOv8. Tunnelling and Underground Space Technology, 158, 106440.
DOI: 10.1016/j.tust.2025.106440
Xie, A. Y., & Li, B. Q. (2024). Transfer learning framework for multi-scale crack type classification with sparse microseismic networks. International Journal of Mining Science and Technology, 34(2), 167–178. DOI: 10.1016/j.ijmst.2024.01.003