Papers
Topics
Authors
Recent
Gemini 2.5 Flash
Gemini 2.5 Flash
139 tokens/sec
GPT-4o
47 tokens/sec
Gemini 2.5 Pro Pro
43 tokens/sec
o3 Pro
4 tokens/sec
GPT-4.1 Pro
47 tokens/sec
DeepSeek R1 via Azure Pro
28 tokens/sec
2000 character limit reached

Predicting machine failures from multivariate time series: an industrial case study (2402.17804v1)

Published 27 Feb 2024 in cs.LG

Abstract: Non-neural Machine Learning (ML) and Deep Learning (DL) models are often used to predict system failures in the context of industrial maintenance. However, only a few researches jointly assess the effect of varying the amount of past data used to make a prediction and the extension in the future of the forecast. This study evaluates the impact of the size of the reading window and of the prediction window on the performances of models trained to forecast failures in three data sets concerning the operation of (1) an industrial wrapping machine working in discrete sessions, (2) an industrial blood refrigerator working continuously, and (3) a nitrogen generator working continuously. The problem is formulated as a binary classification task that assigns the positive label to the prediction window based on the probability of a failure to occur in such an interval. Six algorithms (logistic regression, random forest, support vector machine, LSTM, ConvLSTM, and Transformers) are compared using multivariate telemetry time series. The results indicate that, in the considered scenarios, the dimension of the prediction windows plays a crucial role and highlight the effectiveness of DL approaches at classifying data with diverse time-dependent patterns preceding a failure and the effectiveness of ML approaches at classifying similar and repetitive patterns preceding a failure.

Definition Search Book Streamline Icon: https://streamlinehq.com
References (83)
  1. doi:10.1109/emr.2019.2958037. URL https://doi.org/10.1109/emr.2019.2958037
  2. doi:10.1016/j.cie.2020.106889. URL https://doi.org/10.1016/j.cie.2020.106889
  3. doi:10.1016/j.compind.2020.103298. URL https://doi.org/10.1016/j.compind.2020.103298
  4. doi:10.1080/00207549408957005. URL https://doi.org/10.1080/00207549408957005
  5. doi:10.1016/j.engfailanal.2013.10.006. URL https://doi.org/10.1016/j.engfailanal.2013.10.006
  6. doi:10.1016/0377-2217(95)00313-4. URL https://doi.org/10.1016/0377-2217(95)00313-4
  7. doi:10.1109/tr.2010.2041972. URL https://doi.org/10.1109/tr.2010.2041972
  8. doi:10.1016/j.neucom.2020.10.084. URL https://doi.org/10.1016/j.neucom.2020.10.084
  9. doi:10.1038/s41598-023-31193-8. URL https://doi.org/10.1038/s41598-023-31193-8
  10. doi:10.1145/1670679.1670680. URL https://doi.org/10.1145/1670679.1670680
  11. doi:10.1016/j.ress.2009.10.009. URL https://doi.org/10.1016/j.ress.2009.10.009
  12. doi:10.1108/ijqrm-12-2021-0439. URL https://doi.org/10.1108/ijqrm-12-2021-0439
  13. doi:10.1007/978-3-030-20948-3_11. URL https://doi.org/10.1007/978-3-030-20948-3_11
  14. doi:10.1177/0954405420970517. URL https://doi.org/10.1177/0954405420970517
  15. doi:10.1109/bigdatacongress.2019.00032. URL https://doi.org/10.1109/bigdatacongress.2019.00032
  16. doi:10.1108/ijqrm-04-2019-0131. URL https://doi.org/10.1108/ijqrm-04-2019-0131
  17. doi:10.1109/isie45063.2020.9152441. URL https://doi.org/10.1109/isie45063.2020.9152441
  18. doi:10.1109/bigcomp54360.2022.00053. URL https://doi.org/10.1109/bigcomp54360.2022.00053
  19. doi:10.1002/we.2352. URL https://doi.org/10.1002/we.2352
  20. doi:10.1155/2022/9288452. URL https://doi.org/10.1155/2022/9288452
  21. doi:10.3390/en11071738. URL https://doi.org/10.3390/en11071738
  22. doi:10.1007/978-3-030-20055-8_13. URL https://doi.org/10.1007/978-3-030-20055-8_13
  23. doi:10.1177/1748006x20976817. URL https://doi.org/10.1177/1748006x20976817
  24. doi:10.1109/tste.2011.2163177. URL https://doi.org/10.1109/tste.2011.2163177
  25. doi:10.1016/j.trc.2014.04.013. URL https://doi.org/10.1016/j.trc.2014.04.013
  26. doi:10.1145/2783258.2788611. URL https://doi.org/10.1145/2783258.2788611
  27. doi:10.1145/3292500.3330680. URL https://doi.org/10.1145/3292500.3330680
  28. doi:10.1007/978-3-030-30241-2_50. URL https://doi.org/10.1007/978-3-030-30241-2_50
  29. doi:10.1016/j.jmsy.2021.08.012. URL https://doi.org/10.1016/j.jmsy.2021.08.012
  30. doi:10.1186/s42162-022-00230-7. URL https://doi.org/10.1186/s42162-022-00230-7
  31. doi:10.1109/tii.2016.2641472. URL https://doi.org/10.1109/tii.2016.2641472
  32. doi:10.1080/00401706.2018.1514327. URL https://doi.org/10.1080/00401706.2018.1514327
  33. doi:10.1007/978-3-031-06430-2_56. URL https://doi.org/10.1007/978-3-031-06430-2_56
  34. doi:10.1109/snpd.2007.284. URL https://doi.org/10.1109/snpd.2007.284
  35. doi:10.1109/ssci50451.2021.9659838. URL https://doi.org/10.1109/ssci50451.2021.9659838
  36. doi:10.48550/ARXIV.1709.00845. URL https://arxiv.org/abs/1709.00845
  37. doi:10.1109/bigdata50022.2020.9378419. URL https://doi.org/10.1109/bigdata50022.2020.9378419
  38. doi:10.1016/j.engappai.2015.02.009. URL https://doi.org/10.1016/j.engappai.2015.02.009
  39. doi:10.1109/icphm.2017.7998308. URL https://doi.org/10.1109/icphm.2017.7998308
  40. doi:10.1109/tencon.2018.8650152. URL https://doi.org/10.1109/tencon.2018.8650152
  41. doi:10.1109/ccem.2018.00019. URL https://doi.org/10.1109/ccem.2018.00019
  42. doi:10.23919/acc.2018.8431901. URL https://doi.org/10.23919/acc.2018.8431901
  43. doi:10.1109/tii.2014.2349359. URL https://doi.org/10.1109/tii.2014.2349359
  44. doi:10.3390/su12114776. URL https://doi.org/10.3390/su12114776
  45. doi:10.1162/neco_a_01199. URL https://doi.org/10.1162/neco_a_01199
  46. doi:10.1038/s41598-019-51219-4. URL https://doi.org/10.1038/s41598-019-51219-4
  47. doi:10.1038/s41598-023-44924-8. URL https://doi.org/10.1038/s41598-023-44924-8
  48. doi:10.1007/978-981-99-0063-3_38. URL https://doi.org/10.1007/978-981-99-0063-3_38
  49. doi:10.1109/maes.2021.3053108. URL https://doi.org/10.1109/maes.2021.3053108
  50. doi:10.48550/ARXIV.1506.04214. URL https://arxiv.org/abs/1506.04214
  51. doi:10.1016/j.eswa.2023.120588. URL https://doi.org/10.1016/j.eswa.2023.120588
  52. doi:10.3390/en15103678. URL https://doi.org/10.3390/en15103678
  53. doi:10.48550/ARXIV.2201.07284. URL https://arxiv.org/abs/2201.07284
  54. doi:10.1109/tnsm.2020.3034647. URL https://doi.org/10.1109/tnsm.2020.3034647
  55. doi:10.1016/j.neucom.2022.04.111. URL https://doi.org/10.1016/j.neucom.2022.04.111
  56. doi:10.1016/j.enbuild.2021.111608. URL https://doi.org/10.1016/j.enbuild.2021.111608
  57. doi:10.1109/bmsb58369.2023.10211144. URL https://doi.org/10.1109/bmsb58369.2023.10211144
  58. doi:10.1016/j.chaos.2015.09.002. URL https://doi.org/10.1016/j.chaos.2015.09.002
  59. doi:10.1016/0013-4694(91)90138-t. URL https://doi.org/10.1016/0013-4694(91)90138-t
  60. doi:10.1093/ajcp/94.1.80. URL https://doi.org/10.1093/ajcp/94.1.80
  61. doi:10.1111/trf.14997. URL https://doi.org/10.1111/trf.14997
  62. doi:10.3390/pr4030029. URL https://doi.org/10.3390/pr4030029
  63. doi:10.1145/1541880.1541882. URL https://doi.org/10.1145/1541880.1541882
  64. doi:10.1109/iri.2015.39. URL https://doi.org/10.1109/iri.2015.39
  65. doi:10.1186/s40537-021-00460-8. URL https://doi.org/10.1186/s40537-021-00460-8
  66. doi:10.1093/biomet/73.2.413. URL https://doi.org/10.1093/biomet/73.2.413
  67. L. Breiman, Machine Learning 45 (1) (2001) 5–32. doi:10.1023/a:1010933404324, [link]. URL https://doi.org/10.1023/a:1010933404324
  68. doi:10.1016/s0925-2312(03)00373-4. URL https://doi.org/10.1016/s0925-2312(03)00373-4
  69. doi:10.1109/bigdata47090.2019.9005997. URL https://doi.org/10.1109/bigdata47090.2019.9005997
  70. doi:10.1609/aaai.v37i9.26317. URL https://doi.org/10.1609/aaai.v37i9.26317
  71. doi:10.48550/ARXIV.2108.10566. URL https://arxiv.org/abs/2108.10566
  72. doi:10.1155/2021/5589075. URL https://doi.org/10.1155/2021/5589075
  73. doi:10.3390/rs15041039. URL https://doi.org/10.3390/rs15041039
  74. doi:10.1038/s41598-021-00220-x. URL https://doi.org/10.1038/s41598-021-00220-x
  75. doi:10.1155/2021/4832864. URL https://doi.org/10.1155/2021/4832864
  76. doi:10.1109/icmla.2018.00227. URL https://doi.org/10.1109/icmla.2018.00227
  77. doi:10.1007/s11269-021-02937-w. URL https://doi.org/10.1007/s11269-021-02937-w
  78. doi:10.1007/s42452-021-04421-x. URL https://doi.org/10.1007/s42452-021-04421-x
  79. doi:10.1038/s41598-019-45685-z. URL https://doi.org/10.1038/s41598-019-45685-z
  80. doi:10.2166/hydro.2007.027. URL https://doi.org/10.2166/hydro.2007.027
  81. doi:10.1016/j.bspc.2022.104165. URL https://doi.org/10.1016/j.bspc.2022.104165
  82. doi:10.21203/rs.3.rs-3102074/v1. URL https://doi.org/10.21203/rs.3.rs-3102074/v1
  83. doi:10.3390/app12094221. URL https://doi.org/10.3390/app12094221
Citations (1)
View on Semantic Scholar

Summary

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

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