Resilient Machine Learning in the Nuclear Industry: Crack Detection as a Case Study

Authors: Anita Khadka, Gregory Epiphaniou, Carsten Maple

Abstract:

There is a dramatic surge in the adoption of Machine Learning (ML) techniques in many areas, including the nuclear industry (such as fault diagnosis and fuel management in nuclear power plants), autonomous systems (including self-driving vehicles), space systems (space debris recovery, for example), medical surgery, network intrusion detection, malware detection, to name a few. Artificial Intelligence (AI) has become a part of everyday modern human life. To date, the predominant focus has been developing underpinning ML algorithms that can improve accuracy, while factors such as resiliency and robustness of algorithms have been largely overlooked. If an adversarial attack is able to compromise the learning method or data, the consequences can be fatal, especially but not exclusively in safety-critical applications. In this paper, we present an in-depth analysis of five adversarial attacks and two defence methods on a crack detection ML model. Our analysis shows that it can be dangerous to adopt ML techniques without rigorous testing, since they may be vulnerable to adversarial attacks, especially in security-critical areas such as the nuclear industry. We observed that while the adopted defence methods can effectively defend against different attacks, none of them could protect against all five adversarial attacks entirely.

Keywords: Resilient Machine Learning, attacks, defences, nuclear industry, crack detection.

References:

[1] “Nuclear power reactors,” https://world-nuclear.org/information-library/ nuclear-fuel-cycle/nuclear-power-reactors/nuclear-power-reactors.aspx, 2022.
[2] J. F. Ahearne, A. V. C. Jr, H. A. Feiveson, D. Ingersoll, A. C. Klein, S. Maloney, I. Oelrich, S. Squassoni, and R. Wolfson, “The future of nuclear power in the united states,” 2012.
[3] F.-C. Chen and M. R. Jahanshahi, “Nb-cnn: Deep learning-based crack detection using convolutional neural network and na¨ıve bayes data fusion,” IEEE Transactions on Industrial Electronics, vol. 65, no. 5, pp. 4392–4400, 2018.
[4] Advanced Control Systems to Improve Nuclear Power Plant Reliability and Efficiency, ser. TECDOC Series. Vienna: INTERNATIONAL ATOMIC ENERGY AGENCY, 1997, no. 952.
[Online]. Available: https://www.iaea.org/publications/5604/advanced-control-systems-toimprove- nuclear-power-plant-reliability-and-efficiency
[5] B. Sovacool, “A critical evaluation of nuclear power and renewable electricity in asia,” Journal of Contemporary Asia, vol. 40, pp. 369 – 400, 2010.
[6] S. Suman, “Artificial intelligence in nuclear industry: Chimera or solution?” Journal of Cleaner Production, vol. 278, p. 124022, 2021.
[Online]. Available: https://www.sciencedirect.com/science/article/pii/S0959652620340671
[7] S. Cumblidge, M. T. Anderson, S. Doctor, F. Simonen, and A. Elliot, “An assessment of remote visual methods to detect cracking in reactor components,” 2008.
[8] S. J. Schmugge, L. Rice, N. R. Nguyen, J. Lindberg, R. Grizzi, C. Joffe, and M. C. Shin, “Detection of cracks in nuclear power plant using spatial-temporal grouping of local patches,” in 2016 IEEE Winter Conference on Applications of Computer Vision (WACV), 2016, pp. 1–7.
[9] M. Gavil´an, D. Balcones, O. Marcos, D. F. Llorca, M. A. Sotelo, I. Parra, M. Oca˜na, P. Aliseda, P. Yarza, and A. Am´ırola, “Adaptive road crack detection system by pavement classification,” Sensors, vol. 11, no. 10, pp. 9628–9657, 2011.
[Online]. Available: https://www.mdpi.com/1424-8220/11/10/9628
[10] Y. Sari, P. B. Prakoso, and A. R. Baskara, “Road crack detection using support vector machine (svm) and otsu algorithm,” in 2019 6th International Conference on Electric Vehicular Technology (ICEVT), 2019, pp. 349–354.
[11] Y. Xu, S. Li, D. Zhang, Y. Jin, F. Zhang, N. Li, and H. Li, “Identification framework for cracks on a steel structure surface by a restricted boltzmann machines algorithm based on consumer-grade camera images,” Structural Control and Health Monitoring, vol. 25, no. 2, p. e2075, 2018, e2075 STC-16-0276.R1.
[Online]. Available: https://onlinelibrary.wiley.com/doi/abs/10.1002/stc.2075
[12] L. Zhang, F. Yang, Y. Daniel Zhang, and Y. J. Zhu, “Road crack detection using deep convolutional neural network,” in 2016 IEEE International Conference on Image Processing (ICIP), 2016, pp. 3708–3712.
[13] K. Chen, A. Yadav, A. Khan, Y. Meng, K. Zhu, and Q.-F. Liu, “Improved crack detection and recognition based on convolutional neural network,” Model. Simul. Eng., vol. 2019, Jan. 2019.
[Online]. Available: https://doi.org/10.1155/2019/8796743
[14] S. Li and X. Zhao, “Image-based concrete crack detection using convolutional neural network and exhaustive search technique,” Advances in Civil Engineering, vol. 2019, pp. 1–12, 2019.
[15] K. Gopalakrishnan, H. Gholami, A. Vidyadharan, Alok, Choudhary, and A. Agrawal, “Crack damage detection in unmanned aerial vehicle images of civil infrastructure using pre-trained deep learning model,” 2018.
[16] F. Kucuksubasi and A. G. Sorguc, “Transfer learning-based crack detection by autonomous uavs,” in Proceedings of the 35th International Symposium on Automation and Robotics in Construction (ISARC), J. Teizer, Ed. Taipei, Taiwan: International Association for Automation and Robotics in Construction (IAARC), July 2018, pp. 593–600.
[17] J. J. Kim, A.-R. Kim, and S.-W. Lee, “Artificial neural network-based automated crack detection and analysis for the inspection of concrete structures,” Applied Sciences, vol. 10, no. 22, 2020.
[Online]. Available: https://www.mdpi.com/2076-3417/10/22/8105
[18] Y.-J. Cha, W. Choi, and O. B¨uy¨uk¨ozt¨urk, “Deep learning-based crack damage detection using convolutional neural networks,” Computer-Aided Civil and Infrastructure Engineering, vol. 32, no. 5, pp. 361–378, 2017.
[Online]. Available: https://onlinelibrary.wiley.com/doi/abs/10.1111/mice.12263
[19] N. Papernot, P. McDaniel, A. Sinha, and M. P. Wellman, “Sok: Security and privacy in machine learning,” in 2018 IEEE European Symposium on Security and Privacy (EuroS P), 2018, pp. 399–414.
[20] C. Szegedy, W. Zaremba, I. Sutskever, J. Bruna, D. Erhan, I. Goodfellow, and R. Fergus, “Intriguing properties of neural networks,” 2014.
[21] N. Papernot, P. Mcdaniel, I. Goodfellow, S. Jha, Z. B. Celik, and A. Swami, “Practical black-box attacks against machine learning,” Proceedings of the 2017 ACM on Asia Conference on Computer and Communications Security, 2017.
[22] B. Biggio, B. Nelson, and P. Laskov, “Support vector machines under adversarial label noise,” in Proceedings of the Asian Conference on Machine Learning, ser. Proceedings of Machine Learning Research, C.-N. Hsu and W. S. Lee, Eds., vol. 20. South Garden Hotels and Resorts, Taoyuan, Taiwain: PMLR, 14–15 Nov 2011, pp. 97–112.
[Online]. Available: http://proceedings.mlr.press/v20/biggio11.html
[23] I. J. Goodfellow, J. Shlens, and C. Szegedy, “Explaining and harnessing adversarial examples,” 2015.
[24] I. Rosenberg, A. Shabtai, L. Rokach, and Y. Elovici, “Generic black-box end-to-end attack against state of the art api call based malware classifiers,” 2018.
[25] A. Kurakin, I. Goodfellow, and S. Bengio, “Adversarial machine learning at scale,” 2017.
[26] A. Madry, A. Makelov, L. Schmidt, D. Tsipras, and A. Vladu, “Towards deep learning models resistant to adversarial attacks,” 2019.
[27] N. Carlini and D. Wagner, “Towards evaluating the robustness of neural networks,” 2017.
[28] N. Papernot, F. Faghri, N. Carlini, I. Goodfellow, R. Feinman, A. Kurakin, C. Xie, Y. Sharma, T. Brown, A. Roy, A. Matyasko, V. Behzadan, K. Hambardzumyan, Z. Zhang, Y.-L. Juang, Z. Li, R. Sheatsley, A. Garg, J. Uesato, W. Gierke, Y. Dong, D. Berthelot, P. Hendricks, J. Rauber, R. Long, and P. McDaniel, “Technical report on the cleverhans v2.1.0 adversarial examples library,” 2018.
[29] S.-M. Moosavi-Dezfooli, A. Fawzi, and P. Frossard, “Deepfool: a simple and accurate method to fool deep neural networks,” 2016.
[30] A. Kumar and S. Mehta, “A survey on resilient machine learning,” 2017.
[31] A. Chakraborty, M. Alam, V. Dey, A. Chattopadhyay, and D. Mukhopadhyay, “Adversarial attacks and defences: A survey,” ArXiv, vol. abs/1810.00069, 2018.
[32] N. Akhtar and A. Mian, “Threat of adversarial attacks on deep learning in computer vision: A survey,” 2018.
[33] D. Sgandurra, L. Mu˜noz-Gonz´alez, R. Mohsen, and E. C. Lupu, “Automated dynamic analysis of ransomware: Benefits, limitations and use for detection,” 2016.
[34] H. Kannan, A. Kurakin, and I. Goodfellow, “Adversarial logit pairing,” 2018.
[35] A. Madry, A. Makelov, L. Schmidt, D. Tsipras, and A. Vladu, “Towards deep learning models resistant to adversarial attacks,” 2019.
[36] N. Das, M. Shanbhogue, S.-T. Chen, F. Hohman, L. Chen, M. E. Kounavis, and D. H. Chau, “Keeping the bad guys out: Protecting and vaccinating deep learning with jpeg compression,” 2017.
[37] N. Das, M. Shanbhogue, S.-T. Chen, F. Hohman, S. Li, L. Chen, M. E. Kounavis, and D. H. Chau, “Shield: Fast, practical defense and vaccination for deep learning using jpeg compression,” 2018.
[38] S. Dathathri, S. Zheng, Y. Yue, and R. M. Murray, “Detecting adversarial examples via neural fingerprinting,” 2019.
[Online]. Available: https://openreview.net/forum?id=SJekyhCctQ
[39] C. Lyu, K. Huang, and H.-N. Liang, “A unified gradient regularization family for adversarial examples,” 2015 IEEE International Conference on Data Mining, pp. 301–309, 2015.
[40] U. Shaham, Y. Yamada, and S. N. Negahban, “Understanding adversarial training: Increasing local stability of supervised models through robust optimization,” Neurocomputing, vol. 307, pp. 195–204, 2018.
[41] I. J. Goodfellow, J. Pouget-Abadie, M. Mirza, B. Xu, D. Warde-Farley, S. Ozair, A. Courville, and Y. Bengio, “Generative adversarial networks,” 2014.
[42] H. Lee, S. Han, and J. Lee, “Generative adversarial trainer: Defense to adversarial perturbations with gan,” ArXiv, vol. abs/1705.03387, 2017.
[43] G. Jin, S. Shen, D. Zhang, F. Dai, and Y. Zhang, “Ape-gan: Adversarial perturbation elimination with gan,” in ICASSP 2019 - 2019 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), 2019, pp. 3842–3846.
[44] J. Gao, B. Wang, Z. Lin, W. Xu, and Y. Qi, “Deepcloak: Masking deep neural network models for robustness against adversarial samples,” 2017.
[45] D. Meng and H. Chen, “Magnet: A two-pronged defense against adversarial examples,” in Proceedings of the 2017 ACM SIGSAC Conference on Computer and Communications Security, ser. CCS ’17. New York, NY, USA: Association for Computing Machinery, 2017, p. 135–147.
[Online]. Available: https://doi.org/10.1145/3133956.3134057
[46] C. F. M. Ozgenel and A. G. Sorguc¸, “Performance comparison of pretrained convolutional neural networks on crack detection in buildings,” in Proceedings of the 35th International Symposium on Automation and Robotics in Construction (ISARC), J. Teizer, Ed. Taipei, Taiwan: International Association for Automation and Robotics in Construction (IAARC), July 2018, pp. 693–700.
[47] V. B. Rodrigues, “Concrete-crack-classification-model,” https://github.com/vbrodrigues/Concrete-Crack-Classification-Model, 2018.