Papers
Topics
Authors
Recent
Gemini 2.5 Flash
Gemini 2.5 Flash
169 tokens/sec
GPT-4o
7 tokens/sec
Gemini 2.5 Pro Pro
45 tokens/sec
o3 Pro
4 tokens/sec
GPT-4.1 Pro
38 tokens/sec
DeepSeek R1 via Azure Pro
28 tokens/sec
2000 character limit reached

A probabilistic estimation of remaining useful life from censored time-to-event data (2405.01614v1)

Published 2 May 2024 in cs.LG and cs.AI

Abstract: Predicting the remaining useful life (RUL) of ball bearings plays an important role in predictive maintenance. A common definition of the RUL is the time until a bearing is no longer functional, which we denote as an event, and many data-driven methods have been proposed to predict the RUL. However, few studies have addressed the problem of censored data, where this event of interest is not observed, and simply ignoring these observations can lead to an overestimation of the failure risk. In this paper, we propose a probabilistic estimation of RUL using survival analysis that supports censored data. First, we analyze sensor readings from ball bearings in the frequency domain and annotate when a bearing starts to deteriorate by calculating the Kullback-Leibler (KL) divergence between the probability density function (PDF) of the current process and a reference PDF. Second, we train several survival models on the annotated bearing dataset, capable of predicting the RUL over a finite time horizon using the survival function. This function is guaranteed to be strictly monotonically decreasing and is an intuitive estimation of the remaining lifetime. We demonstrate our approach in the XJTU-SY dataset using cross-validation and find that Random Survival Forests consistently outperforms both non-neural networks and neural networks in terms of the mean absolute error (MAE). Our work encourages the inclusion of censored data in predictive maintenance models and highlights the unique advantages that survival analysis offers when it comes to probabilistic RUL estimation and early fault detection.

Definition Search Book Streamline Icon: https://streamlinehq.com
References (77)
  1. Remaining useful life prediction for ball bearings based on health indicators. MATEC Web of Conferences 261, 02003. doi:10.1051/matecconf/201926102003.
  2. Generalization of deep neural network for bearing fault diagnosis under different working conditions using multiple kernel method. Neurocomputing 352, 42–53. doi:https://doi.org/10.1016/j.neucom.2019.04.010.
  3. Generalised linear models for correlated pseudo‐observations, with applications to multi‐state models. Biometrika 90, 15–27. doi:10.1093/biomet/90.1.15.
  4. Pseudo-observations in survival analysis. Statistical Methods in Medical Research 19, 71–99. doi:10.1177/0962280209105020.
  5. Generating survival times to simulate cox proportional hazards models with time-varying covariates. Statistics in Medicine 31, 3946–3958. doi:https://doi.org/10.1002/sim.5452.
  6. Application of ai tools in fault diagnosis of electrical machines and drives-an overview. IEEE Transactions on Energy Conversion 18, 245–251. doi:https://doi.org/10.1109/TEC.2003.811739.
  7. A classifier fusion system for bearing fault diagnosis. Expert Systems with Applications 40, 6788–6797. doi:https://doi.org/10.1016/j.eswa.2013.06.033.
  8. Allowing for mandatory covariates in boosting estimation of sparse high-dimensional survival models. BMC Bioinformatics 9, 14. doi:10.1186/1471-2105-9-14.
  9. A review on data-driven fault severity assessment in rolling bearings. Mechanical Systems and Signal Processing 99, 169–196. doi:https://doi.org/10.1016/j.ymssp.2017.06.012.
  10. Predictive maintenance using cox proportional hazard deep learning. Advanced Engineering Informatics 44, 101054. doi:https://doi.org/10.1016/j.aei.2020.101054.
  11. A convolutional neural network based degradation indicator construction and health prognosis using bidirectional long short-term memory network for rolling bearings. Advanced Engineering Informatics 48, 101247. doi:https://doi.org/10.1016/j.aei.2021.101247.
  12. On the properties of neural machine translation: Encoder–decoder approaches, in: Proceedings of SSST-8, Eighth Workshop on Syntax, Semantics and Structure in Statistical Translation, pp. 103–111. doi:10.3115/v1/W14-4012.
  13. Regression Models and Life-Tables. Journal of the Royal Statistical Society 34, 187–202.
  14. A review of physics-based models in prognostics: Application to gears and bearings of rotating machinery. Advances in Mechanical Engineering 8, 1687814016664660. doi:https://doi.org/10.1177/1687814016664660.
  15. Kullback-leibler divergence for fault estimation and isolation : Application to gamma distributed data. Mechanical Systems and Signal Processing 93, 118–135. doi:https://doi.org/10.1016/j.ymssp.2017.01.045.
  16. Combining relevance vector machines and exponential regression for bearing residual life estimation. Mechanical Systems and Signal Processing 31, 405–427. doi:https://doi.org/10.1016/j.ymssp.2012.03.011.
  17. Bearing degradation process prediction based on the pca and optimized ls-svm model. Measurement 46, 3143–3152. doi:https://doi.org/10.1016/j.measurement.2013.06.038.
  18. Remaining Useful Life prediction and challenges: A literature review on the use of Machine Learning Methods. Journal of Manufacturing Systems 63, 550–562.
  19. Recent developments of induction motor drives fault diagnosis using ai techniques. IEEE Transactions on Industrial Electronics 47, 994–1004. doi:https://doi.org/10.1109/41.873207.
  20. Greedy Function Approximation: A Gradient Boosting Machine. The Annals of Statistics 29, 1189–1232. doi:10.1214/aos/1013203451.
  21. Dropout as a bayesian approximation: Representing model uncertainty in deep learning, in: Proceedings of The 33rd International Conference on Machine Learning, pp. 1050–1059.
  22. An introduction to statistical learning: with applications in R. 2 ed., Spinger.
  23. Learning to forget: continual prediction with lstm, in: 1999 Ninth International Conference on Artificial Neural Networks ICANN 99. (Conf. Publ. No. 470), pp. 850–855. doi:https://doi.org/10.1049/cp:19991218.
  24. Deep Learning. MIT Press.
  25. A recurrent neural network based health indicator for remaining useful life prediction of bearings. Neurocomputing 240, 98–109. doi:https://doi.org/10.1016/j.neucom.2017.02.045.
  26. Effective ways to build and evaluate individual survival distributions. Journal of Machine Learning Research 21, 1–63.
  27. Survival analysis for HDLSS data with time dependent variables: Lessons from predictive maintenance at a mining service provider, in: Proceedings of 2013 IEEE International Conference on Service Operations and Logistics, and Informatics, pp. 372–381. doi:https://doi.org/10.1109/SOLI.2013.6611443.
  28. Residual life predictions for ball bearings based on self-organizing map and back propagation neural network methods. Mechanical Systems and Signal Processing 21, 193–207. doi:https://doi.org/10.1016/j.ymssp.2005.11.008.
  29. Random Survival Forests. The Annals of Applied Statistics 2, 841–860. doi:https://doi.org/10.1214/08-AOAS169.
  30. ISO, 2007. Rolling Bearings - Dynamic Load rating and rating life.
  31. A neural network constructed by deep learning technique and its application to intelligent fault diagnosis of machines. Neurocomputing 272, 619–628. doi:https://doi.org/10.1016/j.neucom.2017.07.032.
  32. Nonparametric Estimation from Incomplete Observations. Journal of the American Statistical Association 53, 457–481. doi:10.2307/2281868.
  33. Deep learning-based survival prediction of oral cancer patients. Scientific Reports 9, 6994. doi:https://doi.org/10.1038/s41598-019-43372-7.
  34. Prognostics and Health Management of Engineering Systems. Springer International Publishing.
  35. On information and sufficiency. The annals of mathematical statistics 22, 79–86.
  36. Discovering prognostic features using genetic programming in remaining useful life prediction. IEEE Transactions on Industrial Electronics 61, 2464–2472. doi:https://doi.org/10.1109/TIE.2013.2270212.
  37. Perspectives on data-driven models and its potentials in metal forming and blanking technologies. Production Engineering 16, 607–625. doi:https://doi.org/10.1007/s11740-022-01115-0.
  38. Uncertainty estimation in deep bayesian survival models, in: 2023 IEEE EMBS International Conference on Biomedical and Health Informatics (BHI), pp. 1–4. doi:https://doi.org/10.1109/BHI58575.2023.10313466.
  39. Predicting survival time of ball bearings in the presence of censoring. Accepted at AAAI Fall Symposium 2023 on Survival Prediction. arXiv:2309.07188.
  40. Artificial intelligence for fault diagnosis of rotating machinery: A review. Mechanical Systems and Signal Processing 108, 33–47. doi:https://doi.org/10.1016/j.ymssp.2018.02.016.
  41. Bayesian learning for neural networks: an algorithmic survey. Artificial Intelligence Review , 1–51.
  42. Methods for interpreting and understanding deep neural networks. Digital Signal Processing 73, 1–15. doi:https://doi.org/10.1016/j.dsp.2017.10.011.
  43. Deep survival machines: Fully parametric survival regression and representation learning for censored data with competing risks. IEEE Journal of Biomedical and Health Informatics 25, 3163–3175. doi:https://doi.org/10.1109/JBHI.2021.3052441.
  44. Bayesian deep-learning-based health prognostics toward prognostics uncertainty. IEEE Transactions on Industrial Electronics 67, 2283–2293. doi:https://doi.org/10.1109/TIE.2019.2907440.
  45. Chapter 41 estimation of semiparametric models, in: Handbook of Econometrics. volume 4, pp. 2443–2521. doi:https://doi.org/10.1016/S1573-4412(05)80010-8.
  46. An effective meaningful way to evaluate survival models. Proceedings of Machine Learning Research 202, 28244–28276.
  47. Using bayesian neural networks to select features and compute credible intervals for personalized survival prediction. IEEE Transactions on Biomedical Engineering 70, 3389–3400. doi:https://doi.org/10.1109/TBME.2023.3287514.
  48. Cross-domain fault diagnosis of rolling bearing using similar features-based transfer approach. Measurement 172, 108900. doi:https://doi.org/10.1016/j.measurement.2020.108900.
  49. Wavelet filter-based weak signature detection method and its application on rolling element bearing prognostics. Journal of Sound and Vibration 289, 1066–1090. doi:https://doi.org/10.1016/j.jsv.2005.03.007.
  50. Vibration–based Condition Monitoring: Industrial, Automotive and Aerospace Applications. 2 ed., John Wiley & Sons, Ltd.
  51. Rolling element bearing diagnostics—a tutorial. Mechanical Systems and Signal Processing 25, 485–520. doi:https://doi.org/10.1016/j.ymssp.2010.07.017.
  52. Metrics for Offline Evaluation of Prognostic Performance. International Journal of Prognostics and Health Management 1. doi:https://doi.org/10.36001/ijphm.2010.v1i1.1336.
  53. A novel deep learning model for the detection and identification of rolling element-bearing faults. Sensors 20. doi:https://doi.org/10.3390/s20185112.
  54. Extended kalman filtering for remaining-useful-life estimation of bearings. IEEE Transactions on Industrial Electronics 62, 1781–1790. doi:https://doi.org/10.1109/TIE.2014.2336616.
  55. SKF, 2017. Bearing damage and failure analysis. Technical Report. SKF. URL: https://cdn.skfmediahub.skf.com/api/public/0901d1968064c148/pdf_preview_medium/0901d1968064c148_pdf_preview_medium.pdf.
  56. Survival Analysis Methods for Personal Loan Data. Operations Research 50, 277–289.
  57. An end-to-end framework for remaining useful life prediction of rolling bearing based on feature pre-extraction mechanism and deep adaptive transformer model. Computers & Industrial Engineering 161, 107531. doi:https://doi.org/10.1016/j.cie.2021.107531.
  58. Experimental Investigation of the Diagnosis of Angular Contact Ball Bearings Using Acoustic Emission Method and Empirical Mode Decomposition. Advances in Tribology 2020, 1–14. doi:https://doi.org/10.1155/2020/8231752.
  59. Machine Learning: A Bayesian and Optimization Perspective. 2 ed., Elsevier Science.
  60. Rolling bearing fault diagnosis under variable conditions using lmd-svd and extreme learning machine. Mechanism and Machine Theory 90, 175–186. doi:https://doi.org/10.1016/j.mechmachtheory.2015.03.014.
  61. A data-driven failure prognostics method based on mixture of gaussians hidden markov models. IEEE Transactions on Reliability 61, 491–503. doi:https://doi.org/10.1109/TR.2012.2194177.
  62. A hybrid prognostics approach for estimating remaining useful life of rolling element bearings. IEEE Transactions on Reliability 69, 401–412. doi:https://doi.org/10.1109/TR.2018.2882682.
  63. Predictive maintenance based on event-log analysis: A case study. IBM Journal of Research and Development 61, 11:121–11:132. doi:https://doi.org/10.1147/JRD.2017.2648298.
  64. A bayesian inference-based approach for performance prognostics towards uncertainty quantification and its applications on the marine diesel engine. ISA Transactions 118, 159–173. doi:https://doi.org/10.1016/j.isatra.2021.02.024.
  65. Remaining useful life prediction of rolling bearings based on pearson correlation-kpca multi-feature fusion. Measurement 201, 111572. doi:https://doi.org/10.1016/j.measurement.2022.111572.
  66. Research on a remaining useful life prediction method for degradation angle identification two-stage degradation process. Mechanical Systems and Signal Processing 184, 109747. doi:https://doi.org/10.1016/j.ymssp.2022.109747.
  67. Application of relevance vector machine and survival probability to machine degradation assessment. Expert Systems with Applications 38, 2592–2599. doi:https://doi.org/10.1016/j.eswa.2010.08.049.
  68. Machine health prognostics using survival probability and support vector machine. Expert Systems with Applications 38, 8430–8437. doi:https://doi.org/10.1016/j.eswa.2011.01.038.
  69. Improved deep pca and kullback–leibler divergence based incipient fault detection and isolation of high-speed railway traction devices. Sustainable Energy Technologies and Assessments 57, 103208. doi:https://doi.org/10.1016/j.seta.2023.103208.
  70. Rul prediction for rolling bearings based on convolutional autoencoder and status degradation model. Applied Soft Computing 130, 109686. doi:https://doi.org/10.1016/j.asoc.2022.109686.
  71. A novel health indicator for intelligent prediction of rolling bearing remaining useful life based on unsupervised learning model. Computers & Industrial Engineering 176, 108999. doi:https://doi.org/10.1016/j.cie.2023.108999.
  72. Learning patient-specific cancer survival distributions as a sequence of dependent regressors, in: Advances in Neural Information Processing Systems, pp. 1845––1853.
  73. Detecting abnormal situations using the kullback–leibler divergence. Automatica 50, 2777–2786. doi:https://doi.org/10.1016/j.automatica.2014.09.005.
  74. Multiobjective deep belief networks ensemble for remaining useful life estimation in prognostics. IEEE Transactions on Neural Networks and Learning Systems 28, 2306–2318. doi:https://doi.org/10.1109/TNNLS.2016.2582798.
  75. A survey of condition monitoring and protection methods for medium-voltage induction motors. IEEE Transactions on Industry Applications 47, 34–46. doi:https://doi.org/10.1109/TIA.2010.2090839.
  76. A data-driven approach for remaining useful life prediction of aircraft engines, in: 2018 21st International Conference on Intelligent Transportation Systems (ITSC), pp. 184–189. doi:https://doi.org/10.1109/ITSC.2018.8569915.
  77. Deep convolutional neural network for survival analysis with pathological images, in: 2016 IEEE International Conference on Bioinformatics and Biomedicine (BIBM), pp. 544–547. doi:https://doi.org/10.1109/BIBM.2016.7822579.

Summary

We haven't generated a summary for this paper yet.

Ai Sparkles Streamline Icon: https://streamlinehq.com Summarize Now
X Twitter Logo Streamline Icon: https://streamlinehq.com

Tweets