Explainable AI for Machine Fault Diagnosis: Understanding Features’ Contribution in Machine Learning Models for Industrial Condition Monitoring
Source abstract
Abstract
Although the effectiveness of machine learning (ML) for machine diagnosis has been widely established, the interpretation of the diagnosis outcomes is still an open issue. Machine learning models behave as black boxes; therefore, the contribution given by each of the selected features to the diagnosis is not transparent to the user. This work is aimed at investigating the capabilities of the SHapley Additive exPlanation (SHAP) to identify the most important features for fault detection and classification in condition monitoring programs for rotating machinery. The authors analyse the case of medium-sized bearings of industrial interest. Namely, vibration data were collected for different health states from the test rig for industrial bearings available at the Mechanical Engineering Laboratory of Politecnico di Torino. The Support Vector Machine (SVM) and k-Nearest Neighbour (kNN) diagnosis models are explained by means of the SHAP. Accuracies higher than 98.5% are achieved for both the models using the SHAP as a criterion for feature selection. It is found that the skewness and the shape factor of the vibration signal have the greatest impact on the models’ outcomes.
Source: crossref · Quality: baseline
Research context
OpenAlex topics: Machine Fault Diagnosis Techniques · Fault Detection and Control Systems · Engineering Diagnostics and Reliability
SciVX discovery tags are terminology-based. OpenAlex topics are source-provided classifications.
Affiliations
Dipartimento di Ingegneria Meccanica e Aerospaziale (DIMEAS), Politecnico di Torino, Corso Duca Degli Abruzzi 24, 10129 Torino, Italy
References (100+)
Show source-provided references
1. A New Approach to Intelligent Fault Diagnosis of Rotating Machinery10.1016/j.eswa.2007.08.072 ↗
2. Gupta, P.K. (1984). Advanced Dynamics of Rolling Elements, Springer.10.1007/978-1-4612-5276-4 ↗
3. An Extensive Review of Vibration Modelling of Rolling Element Bearings with Localised and Extended Defects10.1016/j.jsv.2015.04.037 ↗
4. A Novel Optimized SVM Classification Algorithm with Multi-Domain Feature and Its Application to Fault Diagnosis of Rolling Bearing10.1016/j.neucom.2018.05.002 ↗
5. Hasan, M.J., Sohaib, M., and Kim, J.-M. (2020). A Multitask-Aided Transfer Learning-Based Diagnostic Framework for Bearings under Inconsistent Working Conditions. Sensors, 20.10.3390/s20247205 ↗
6. Quantitative and Localization Diagnosis of a Defective Ball Bearing Based on Vertical–Horizontal Synchronization Signal Analysis10.1109/TIE.2017.2698359 ↗
7. Fault Diagnosis of Intershaft Bearings Using Fusion Information Exergy Distance Method10.1155/2018/7546128 ↗
8. A Novel Health Indicator Based on the Lyapunov Exponent, a Probabilistic Self-Organizing Map, and the Gini-Simpson Index for Calculating the RUL of Bearings10.1016/j.measurement.2020.108002 ↗
9. Hasan, M.J., Sohaib, M., and Kim, J.-M. (2021). An Explainable Ai-Based Fault Diagnosis Model for Bearings. Sensors, 21.10.3390/s21124070 ↗
10. An Overview of Bearing Vibration Analysis
11. Intelligent Fault Identification Based on Multisource Domain Generalization Towards Actual Diagnosis Scenario10.1109/TIE.2019.2898619 ↗
12. Scalable and Unsupervised Feature Engineering Using Vibration-Imaging and Deep Learning for Rotor System Diagnosis10.1109/TIE.2017.2752151 ↗
13. Brusa, E., Bruzzone, F., Delprete, C., Di Maggio, L.G., and Rosso, C. (2020). Health Indicators Construction for Damage Level Assessment in Bearing Diagnostics: A Proposal of an Energetic Approach Based on Envelope Analysis. Appl. Sci., 10.10.3390/app10228131 ↗
14. Bearing Health Monitoring Based on the Orthogonal Empirical Mode Decomposition
15. Rolling Bearings Monitoring and Damage Detection Methodology10.4028/www.scientific.net/AMM.3-4.293 ↗
16. Brusa, E., Delprete, C., and Giorio, L. (2022). Smart Manufacturing in Rolling Process Based on Thermal Safety Monitoring by Fiber Optics Sensors Equipping Mill Bearings. Appl. Sci., 12.10.3390/app12094186 ↗
17. Refined Composite Multiscale Fuzzy Entropy: Localized Defect Detection of Rolling Element Bearing10.1007/s12206-018-1211-8 ↗
18. Fault Diagnosis of Rotation Machinery Based on Support Vector Machine Optimized by Quantum Genetic Algorithm10.1109/ACCESS.2018.2789933 ↗
19. Reliable Fault Diagnosis for Low-Speed Bearings Using Individually Trained Support Vector Machines With Kernel Discriminative Feature Analysis10.1109/TPEL.2014.2358494 ↗
20. Fault Diagnosis of Low Speed Bearing Based on Relevance Vector Machine and Support Vector Machine10.1016/j.eswa.2008.09.033 ↗
21. Brusa, E., Delprete, C., and Di Maggio, L.G. (2022). Eigen-Spectrograms: An Interpretable Feature Space for Bearing Fault Diagnosis Based on Artificial Intelligence and Image Processing. Mech. Adv. Mater. Struct., 1–13.10.1080/15376494.2022.2102274 ↗
22. Plastic Bearing Fault Diagnosis Based on a Two-Step Data Mining Approach
23. Using Multi-Sensor Data Fusion for Vibration Fault Diagnosis of Rolling Element Bearings by Accelerometer and Load Cell10.1016/j.inffus.2013.10.002 ↗
24. THIRD-ORDER SPECTRAL TECHNIQUES FOR THE DIAGNOSIS OF MOTOR BEARING CONDITION USING ARTIFICIAL NEURAL NETWORKS10.1006/mssp.2001.1469 ↗
25. Vibration Analysis for Bearing Fault Detection and Classification Using an Intelligent Filter10.1016/j.mechatronics.2014.01.003 ↗
26. A Novel Bearing Fault Diagnosis Method Based on Gaussian Restricted Boltzmann Machine10.1155/2016/2957083 ↗
27. Guo, S., Yang, T., Gao, W., and Zhang, C. (2018). A Novel Fault Diagnosis Method for Rotating Machinery Based on a Convolutional Neural Network. Sensors, 18.10.3390/s18051429 ↗
28. Brusa, E., Delprete, C., and Di Maggio, L.G. (2021). Deep Transfer Learning for Machine Diagnosis: From Sound and Music Recognition to Bearing Fault Detection. Appl. Sci., 11.10.3390/app112411663 ↗
29. Di Maggio, L.G. (2022). Intelligent Fault Diagnosis of Industrial Bearings Using Transfer Learning and CNNs Pre-Trained for Audio Classification. Sensors, 23.10.3390/s23010211 ↗
30. Applications of Machine Learning to Machine Fault Diagnosis: A Review and Roadmap10.1016/j.ymssp.2019.106587 ↗
31. Islam, M.R., Islam, M.M.M., and Kim, J.-M. (2016, January 20–22). Feature Selection Techniques for Increasing Reliability of Fault Diagnosis of Bearings. Proceedings of the 2016 9th International Conference on Electrical and Computer Engineering (ICECE), Dhaka, Bangladesh.10.1109/ICECE.2016.7853940 ↗
32. Roelofs, R., Shankar, V., Recht, B., Fridovich-Keil, S., Hardt, M., Miller, J., and Schmidt, L. (2019). Proceedings of the Advances in Neural Information Processing Systems, Curran Associates, Inc.
33. Yassine, A., Mohamed, C., and Zinedine, A. (2017, January 17–19). Feature Selection Based on Pairwise Evalution. Proceedings of the 2017 Intelligent Systems and Computer Vision (ISCV), Fez, Morocco.10.1109/ISACV.2017.8054919 ↗
34. Shapley Additive Explanations for NO2 Forecasting10.1016/j.ecoinf.2019.101039 ↗
35. Bergadano, F., and De Raedt, L. (1994). Proceedings of the Machine Learning: ECML-94, Springer.10.1007/3-540-57868-4 ↗
36. Hall, M.A., and Smith, L.A. (1998, January 6). Practical Feature Subset Selection for Machine Learning. Proceedings of the 21st Australasian Computer Science Conference ACSC’98, Perth, Australia.
37. Feature Selection Based on Mutual Information Criteria of Max-Dependency, Max-Relevance, and Min-Redundancy10.1109/TPAMI.2005.159 ↗
38. Bishop, C.M. (1995). Neural Networks for Pattern Recognition, Oxford University Press.10.1201/9781420050646.ptb6 ↗
39. ART–KOHONEN Neural Network for Fault Diagnosis of Rotating Machinery10.1016/S0888-3270(03)00073-6 ↗
40. Landau, S., Leese, M., Stahl, D., and Everitt, B.S. (2011). Cluster Analysis, John Wiley & Sons.
41. Hui, K.H., Ooi, C., Lim, M., Leong, M., and Al-Obaidi, S. (2017). An Improved Wrapper-Based Feature Selection Method for Machinery Fault Diagnosis. PLoS ONE, 12.10.1371/journal.pone.0189143 ↗
42. Sparsity and Smoothness via the Fused Lasso10.1111/j.1467-9868.2005.00490.x ↗
43. Tikhonov, A.N., Goncharsky, A.V., Stepanov, V.V., and Yagola, A.G. (1995). Numerical Methods for the Solution of Ill-Posed Problems, Springer.10.1007/978-94-015-8480-7 ↗
44. Survey on Data-Driven Industrial Process Monitoring and Diagnosis10.1016/j.arcontrol.2012.09.004 ↗
45. A Novel Transmission Line Relaying Scheme for Fault Detection and Classification Using Wavelet Transform and Linear Discriminant Analysis10.1016/j.asej.2014.10.005 ↗
46. Process Monitoring and Diagnosis by Multiblock PLS Methods10.1002/aic.690400509 ↗
47. Guyon, I., Luxburg, U.V., Bengio, S., Wallach, H., Fergus, R., Vishwanathan, S., and Garnett, R. (2017). Proceedings of the Advances in Neural Information Processing Systems, Curran Associates, Inc.
48. Choi, S.H., and Lee, J.M. (2022, January 7–9). Explainable Fault Diagnosis Model Using Stacked Autoencoder and Kernel SHAP. Proceedings of the 2022 IEEE International Symposium on Advanced Control of Industrial Processes (AdCONIP), Vancouver, BC, Canada.10.1109/AdCONIP55568.2022.9894124 ↗
49. An Explainable Artificial Intelligence Approach for Unsupervised Fault Detection and Diagnosis in Rotating Machinery10.1016/j.ymssp.2021.108105 ↗
50. Peeking Inside the Black-Box: A Survey on Explainable Artificial Intelligence (XAI)10.1109/ACCESS.2018.2870052 ↗
51. Doran, D., Schulz, S., and Besold, T.R. (2017). What Does Explainable AI Really Mean? A New Conceptualization of Perspectives. arXiv.
52. Shapley Effects for Global Sensitivity Analysis: Theory and Computation10.1137/15M1048070 ↗
53. Explaining Individual Predictions When Features Are Dependent: More Accurate Approximations to Shapley Values10.1016/j.artint.2021.103502 ↗
54. Lundberg, S., Erion, G., Chen, H., DeGrave, A., Prutkin, J., Nair, B., Katz, R., Himmelfarb, J., Bansal, N., and Lee, S.-I. (2019). Explainable AI for Trees: From Local Explanations to Global Understanding. arXiv.
55. Explaining Prediction Models and Individual Predictions with Feature Contributions10.1007/s10115-013-0679-x ↗
56. Explaining Predictive Models with Mixed Features Using Shapley Values and Conditional Inference Trees10.1007/978-3-030-57321-8_7 ↗
57. Moehle, N., Boyd, S., and Ang, A. (2021). Portfolio Performance Attribution via Shapley Value. arXiv.
58. Shapley Values for Feature Selection: The Good, the Bad, and the Axioms10.1109/ACCESS.2021.3119110 ↗
59. An Efficient Explanation of Individual Classifications Using Game Theory
60. Causal Connections between Socioeconomic Disparities and COVID-19 in the USA10.1038/s41598-022-18725-4 ↗
61. Improving Interpretation of Sea-Level Projections through a Machine-Learning-Based Local Explanation Approach10.5194/tc-16-4637-2022 ↗
62. Interpretable Machine Learning for Genomics10.1007/s00439-021-02387-9 ↗
63. A Decision Support System for Safer Airplane Landings: Predicting Runway Conditions Using XGBoost and Explainable AI10.1016/j.coldregions.2022.103556 ↗
64. Dong, H., Sun, J., and Sun, X. (2021). A Multi-Objective Multi-Label Feature Selection Algorithm Based on Shapley Value. Entropy, 23.10.3390/e23081094 ↗
65. Goštautaitė, D., and Sakalauskas, L. (2022). Multi-Label Classification and Explanation Methods for Students’ Learning Style Prediction and Interpretation. Appl. Sci., 12.10.3390/app12115396 ↗
66. Understanding Machine Learning Classifier Decisions in Automated Radiotherapy Quality Assurance10.1088/1361-6560/ac3e0e ↗
67. Association Between Perioperative Adverse Cardiac Events and Mortality During One-Year Follow-Up After Noncardiac Surgery10.1161/JAHA.121.024325 ↗
68. Akimoto, S., Lebreton, P., Takahashi, S., and Yamagishi, K. (2022, January 26–28). Quantitative Causality Analysis of Viewing Abandonment Reasons Using Shapley Value. Proceedings of the 2022 IEEE 24th International Workshop on Multimedia Signal Processing (MMSP), Shanghai, China.10.1109/MMSP55362.2022.9950013 ↗
69. Li, L., Wu, X., Kong, M., Zhou, D., and Tao, X. (2022, January 23–29). Towards the Quantitative Interpretability Analysis of Citizens Happiness Prediction. Proceedings of the 39th International Joint Conference on Artificial Intelligence (IJCAI-ECAI 2022), Vienna, Austria.10.24963/ijcai.2022/707 ↗
70. Investigation on Mechanical Properties and Energy Absorption Capabilities of AlSi10Mg Triply Periodic Minimal Surface Sheet Structures Fabricated via Selective Laser Melting10.1007/s11665-022-06883-5 ↗
71. Remman, S.B., Strumke, I., and Lekkas, A.M. (2022, January 8–10). Causal versus Marginal Shapley Values for Robotic Lever Manipulation Controlled Using Deep Reinforcement Learning. Proceedings of the 2022 American Control Conference (ACC), Atlanta, GA, USA.10.23919/ACC53348.2022.9867807 ↗
72. Cohen, S., Ruppin, E., and Dror, G. (2005, January 19–27). Feature Selection Based on the Shapley Value. Proceedings of the 19th International Joint Conference on Artificial Intelligence, Montreal, QC, Canada.
73. Marcílio, W.E., and Eler, D.M. (2020, January 7–10). From Explanations to Feature Selection: Assessing SHAP Values as Feature Selection Mechanism. Proceedings of the 2020 33rd SIBGRAPI Conference on Graphics, Patterns and Images (SIBGRAPI), Porto de Galinhas, Brazil.10.1109/SIBGRAPI51738.2020.00053 ↗
74. Designing a Feature Selection Method Based on Explainable Artificial Intelligence10.1007/s12525-022-00608-1 ↗
75. CGA: A New Feature Selection Model for Visual Human Action Recognition10.1007/s00521-020-05297-5 ↗
76. Predicting Generalized Anxiety Disorder among Women Using Shapley Value10.1016/j.jiph.2020.02.042 ↗
77. Tripathi, S., Hemachandra, N., and Trivedi, P. (2020, January 10–13). Interpretable Feature Subset Selection: A Shapley Value Based Approach. Proceedings of the 2020 IEEE International Conference on Big Data (Big Data), Atlanta, GA, USA.10.1109/BigData50022.2020.9378102 ↗
78. Mey, O., and Neufeld, D. (2022). Explainable AI Algorithms for Vibration Data-Based Fault Detection: Use Case-Adadpted Methods and Critical Evaluation. Sensors, 22.10.3390/s22239037 ↗
79. Yang, D., Karimi, H.R., and Gelman, L. (2022). A Fuzzy Fusion Rotating Machinery Fault Diagnosis Framework Based on the Enhancement Deep Convolutional Neural Networks. Sensors, 22.10.3390/s22020671 ↗
80. Brusa, E., Delprete, C., Giorio, L., Di Maggio, L.G., and Zanella, V. (2022). Design of an Innovative Test Rig for Industrial Bearing Monitoring with Self-Balancing Layout. Machines, 10.10.3390/machines10010054 ↗
81. Support-Vector Networks10.1007/BF00994018 ↗
82. Support Vector Machine in Machine Condition Monitoring and Fault Diagnosis10.1016/j.ymssp.2006.12.007 ↗
83. Boser, B., Guyon, I., and Vapnik, V. (1992, January 27–29). A Training Algorithm for Optimal Margin Classifier. Proceedings of the Fifth Annual Workshop on Computational Learning Theory, Pittsburgh, PA, USA.10.1145/130385.130401 ↗
84. Bearing Fault Diagnosis and Prognosis Using Data Fusion Based Feature Extraction and Feature Selection10.1016/j.measurement.2021.110506 ↗
85. Vapnik, V. (1999). The Nature of Statistical Learning Theory, Springer Science & Business Media.10.1007/978-1-4757-3264-1 ↗
86. Shalev-Shwartz, S., and Ben-David, S. (2014). Understanding Machine Learning: From Theory to Algorithms, Cambridge University Press. [1st ed.].10.1017/CBO9781107298019 ↗
87. Fix, E. (1985). Discriminatory Analysis: Nonparametric Discrimination, Consistency Properties, USAF School of Aviation Medicine.
88. Hart, P.E., Stork, D.G., and Duda, R.O. (2000). Pattern Classification, Wiley.
89. Nearest Neighbor Pattern Classification10.1109/TIT.1967.1053964 ↗
90. Neighborhood Size Selection in the K-Nearest-Neighbor Rule Using Statistical Confidence10.1016/j.patcog.2005.08.009 ↗
91. The Condensed Nearest Neighbor Rule (Corresp.)10.1109/TIT.1968.1054155 ↗
92. Shapley, L.S. (1997). Classics in Game Theory, The Rand Corporation.
93. Understanding Global Feature Contributions with Additive Importance Measures
94. Daumé, H., and Singh, A. Proceedings of the Proceedings of the 37th International Conference on Machine Learning.
95. An Intelligent Fault Diagnosis Method Using Unsupervised Feature Learning Towards Mechanical Big Data10.1109/TIE.2016.2519325 ↗
96. Yousefi, S. (2022). A Data-Driven Approach for Fault Classification of a Manufacturing Process. [Master’s Thesis, NTNU].
97. A Multidimensional Hybrid Intelligent Method for Gear Fault Diagnosis10.1016/j.eswa.2009.06.060 ↗
98. Caesarendra, W., and Tjahjowidodo, T. (2017). A Review of Feature Extraction Methods in Vibration-Based Condition Monitoring and Its Application for Degradation Trend Estimation of Low-Speed Slew Bearing. Machines, 5.10.3390/machines5040021 ↗
99. Elliott, D.F., and Rao, K.R. (1982). Fast Transforms: Algorithms, Analyses, Applications, Academic Press.
100. Oppenheim, A., and Schafer, R. (2009). Discrete-Time Signal Processing, Pearson. [3rd ed.].