Detection and Analysis of Out-of-Control Observations in Statistical Process Control Using Explainable Artificial Intelligence Models

Abstract

Statistical process control (SPC) and anomaly detection are critical for enhancing product quality and operational efficiency in industrial manufacturing processes. However, traditional multivariate SPC methods cannot be directly applied to data with constant sum constraints, such as CoDa. In this study, the CoDa data obtained from the casting process were transformed into Euclidean space using the isometric log-ratio (ilr) transformation and monitored using the Hotelling T² control chart. Machine learning and explainability methods were employed to detect and understand the root causes of out-of-control signals. In this context, five classification models were compared: SVM, RF, XGBoost, logistic regression, and KNN. The highest test accuracy rate of 93.88% was achieved using the SVM model. To explain the decision mechanism of the model, SHapley Additive exPlanations (SHAP) and the Mason–Young–Tracy (MYT) generalization approach were jointly applied. The findings reveal that the SHAP and MYT results demonstrate a low level of consistency and that the model provides reliable local and global explainability outputs. By overcoming the limitations of traditional SPC methods, this integrated approach facilitates the understanding of root causes of anomalies in the casting process.

Keywords

• Statistical Process Control (SPC)
• Anomaly Detection
• Explainable Artificial Intelligence (XAI)

References

1. Aitchison, J. (1986). The statistical analysis of compositional data. Chapman and Hall. https://doi.org/10.1007/978-94-009-4109-0 google scholar
2. Aitchison, J. (2011). The statistical analysis of compositional data. Springer Netherlands. https://doi.org/10.1007/978-94-009-4109-0 google scholar
3. Arrieta, A. B., Díaz-Rodríguez, N., Del Ser, J., Bennetot, A., Tabik, S., Barbado, A., Garcia, S., Gil-Lopez, S., Molina, D., Benjamins, R., Chatila, R., & Herrera, F. (2020). Explainable artificial intelligence (XAI): Concepts, taxonomies, opportunities and challenges toward responsible AI. Information Fusion, 58, 82–115. https://doi.org/10.1016/j.inffus.2019.12.012 google scholar
4. Baykoç, Ö. F., & Sakallı, Ü. S. (2009). An aggregate production planning model for brass casting industry in fuzzy environment. International Journal of Mathematical and Statistical Sciences, 1(3), 154–158. google scholar
5. Boser, B. E., Guyon, I. M., & Vapnik, V. N. (1992, July). A training algorithm for optimal margin classifiers. In Proceedings of the Fifth Annual Workshop on Computational Learning Theory (pp. 144–152). ACM. google scholar
6. Boyles, R. A. (1997). Using the chi-square statistic to monitor compositional process data. Journal of Applied Statistics, 24(5), 589–602. https://doi.org/10.1080/02664769723567 google scholar
7. Breiman, L. (1996). Bagging predictors. Machine Learning, 24(2), 123–140. google scholar
8. Breiman, L. (2001). Random forests. Machine Learning, 45(1), 5–32. google scholar

9. Chen, T., & Guestrin, C. (2016). XGBoost: A scalable tree boosting system. In Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining (pp. 785–794). ACM. https://doi.org/10.1145/2939672.2939785 google scholar
10. Cobb, B. R., & Li, L. (2019). Bayesian network model for quality control with categorical attribute data. Applied Soft Computing, 84, 1–16. https://doi.org/10.1016/j.asoc.2019.105746 google scholar
11. Cortes, C., & Vapnik, V. (1995). Support-vector networks. Machine Learning, 20, 273–297. https://doi.org/10.1007/BF00994018 google scholar
12. Cutler, A., Cutler, D. R., & Stevens, J. R. (2012). Random forests. In Z. H. Zhou (Ed.), Ensemble machine learning (pp. 157–175). Springer. google scholar
13. Çalıgülü, U., Bölükbaş, D., & Darcan, N. (2023). Savunma sanayinde kullanılan CuZn30 ve CuZn10 pirinç alaşımlarına uygulanan farklı ısıl işlem parametrelerinin mikroyapı ve mekanik özelliklere etkisi. International Journal of Physical and Applied Sciences, 9(2), 402–410. https://doi.org/10.29132/ijpas.1395901 google scholar
14. Çınar, Z. M., Nuhu, A. A., Zeeshan, Q., Korhan, O., Asmael, M., & Safaei, B. (2020). Machine learning in predictive maintenance towards sustainable smart manufacturing in Industry 4.0. Sustainability, 12(19), 8211. https://doi.org/10.3390/su12198211 google scholar
15. Dilki, G., & Başar, Ö. D. (2020). İşletmelerin iflas tahmininde K-en yakın komşu algoritması üzerinden uzaklık ölçütlerinin karşılaştırılması. İstanbul Ticaret Üniversitesi Fen Bilimleri Dergisi, 19(38), 224–233. google scholar
16. Gunning, D., Stefik, M., Choi, J., Miller, T., Stumpf, S., & Yang, G. Z. (2021). XAI—Explainable artificial intelligence. Science Robotics, 6(37), eabg7890. https://doi.org/10.1126/scirobotics.aay7120 google scholar
17. Hotelling, H. (1947). Multivariate quality control illustrated by air testing of sample bombsights. In C. Eisenhart, M. W. Hastay, & W. A. Wallis (Eds.), Techniques of statistical analysis (pp. 111–184). McGraw-Hill. google scholar
18. Imran, M., Sun, J. S., Zaidi, F. S., Abbas, Z., & Nazir, H. Z. (2022a). Multivariate cumulative sum control chart for compositional data with known and estimated process parameters. Quality and Reliability Engineering International, 38(5), 2691–2714. https://doi.org/10.1002/qre.3099 google scholar
19. Imran, M., Sun, J., Zaidi, F. S., Abbas, Z., & Nazir, H. Z. (2022b). On designing efficient multivariate exponentially weighted moving average control chart for compositional data using variable sample size. Journal of Statistical Computation and Simulation, 93(10), 1622–1643. https://doi.org/10.1080/00949655.2022.2146115 google scholar
20. Imran, M., Sun, J., Zaidi, F. S., Abbas, Z., & Nazir, H. Z. (2023a). Evaluating the performance of variable sampling interval Hotelling T² charting scheme for compositional data in the presence of measurement error. Quality and Reliability Engineering International, 39(6), 2125–2151. https://doi.org/10.1002/qre.3307 google scholar
21. Imran, M., Sun, J., Zaidi, F. S., Abbas, Z., & Nazir, H. Z. (2023b). Effect of measurement error on the multivariate CUSUM control chart for compositional data. CMES—Computer Modeling in Engineering and Sciences, 136(2), 1207–1257. https://doi.org/10.32604/cmes.2023.025492 google scholar
22. Imran, M., Sun, J., Hu, X., Zaidi, F. S., & Tang, A. (2023c). Investigating zero-state and steady-state performance of MEWMA-CoDa control chart using variable sampling interval. Journal of Applied Statistics, 51(5), 913–934. https://doi.org/10.1080/02664763.2023.2170336 google scholar
23. Liaw, A., & Wiener, M. (2002). Classification and regression by randomForest. R News, 2(3), 18–22. google scholar
24. Lundberg, S. M., & Lee, S.-I. (2017). A unified approach to interpreting model predictions. In Proceedings of the 31st International Conference on Neural Information Processing Systems (pp. 4766–4777). google scholar
25. Lundberg, S. M., Erion, G. G., & Lee, S.-I. (2018). Consistent individualized feature attribution for tree ensembles. arXiv preprint, arXiv:1802.03888. https://arxiv.org/abs/1802.03888 google scholar
26. Mihirette, S., & Tan, Q. (2022). SHAP algorithm for healthcare data classification. In P. García Bringas, H. Pérez García, F. J. Martínez de Pisón, F. Martínez Álvarez, A. Troncoso Lora, A. Herrero, J. R. Villar, E. A. de la Cal, H. Quintián, & E. Corchado (Eds.), Hybrid artificial intelligence systems (pp. 363–374). Springer International Publishing. https://doi.org/10.1007/978-3-031-15471-3_32 google scholar
27. Montgomery, D. C. (2020). Introduction to statistical quality control (8th ed.). John Wiley & Sons. google scholar
28. Müller, K. R., Mika, S., Tsuda, K., & Schölkopf, K. (2018). An introduction to kernel-based learning algorithms. In Handbook of neural network signal processing (pp. 4–1). CRC Press. google scholar
29. Orçanlı, K. (2021). Döküm sanayinde süreç tabanlı temel gösterimleri ile istatistiksel süreç kontrolü. Academic Platform Journal of Engineering and Science, 9(1), 134–158. https://doi.org/10.21541/apjes.720051 google scholar
30. Orçanlı, K., Birgören, B., & Oktay, E. (2018). Döküm sanayisinde metal alaşım oranlarına Hotelling T² ve MEWMA kontrol grafikleri uygulamaları. Social Sciences Research Journal, 7(1), 114–135. google scholar
31. Özel, S. (2005). Çok değişkenli kalite kontrolün döküm sanayinde uygulanması (Yayınlanmış yüksek lisans tezi). Kırıkkale Üniversitesi, Fen Bilimleri Enstitüsü. google scholar
32. Pawlowsky-Glahn, V., Egozcue, J. J., & Tolosana-Delgado, R. (2015). Modeling and analysis of compositional data. John Wiley & Sons. google scholar
33. Rudin, C. (2019). Stop explaining black box machine learning models for high stakes decisions and use interpretable models instead. Nature Machine Intelligence, 1(5), 206–215. https://doi.org/10.1038/s42256-019-0048-x google scholar
34. Sakalli, U. S., & Birgoren, B. (2008). A spreadsheet-based decision support tool for blending problems in brass casting industry. Computers & Industrial Engineering, 56(2), 724–735. https://doi.org/10.1016/j.cie.2008.05.009 google scholar
35. Shapley, L. S. (1953). A value for n-person games. In H. W. Kuhn & A. W. Tucker (Eds.), Contributions to the theory of games (Vol. 2, pp. 307–318). Princeton University Press. https://doi.org/10.1515/9781400881970-018 google scholar
36. Susto, G. A., Schirru, A., Pampuri, S., McLoone, S., & Beghi, A. (2015). Machine learning for predictive maintenance: A multiple classifier approach. IEEE Transactions on Industrial Informatics, 11(3), 812–820. https://doi.org/10.1109/TII.2014.2349359 google scholar
37. Tran, K. P., Castagliola, P., Celano, G., & Khoo, M. B. C. (2018). Monitoring compositional data using multivariate exponentially weighted moving average scheme. Quality and Reliability Engineering International, 34(3), 391–402. https://doi.org/10.1002/qre.2260 google scholar
38. Vapnik, V. (2013). The nature of statistical learning theory. Springer Science & Business Media. google scholar
39. Vives-Mestres, M., Daunis-i-Estadella, P., & Martín-Fernández, J. A. (2014a). Out-of-control signals in three-part compositional T² control chart. Quality and Reliability Engineering International, 30(3), 337–346. https://doi.org/10.1002/qre.1583 google scholar
40. Vives-Mestres, M., Daunis-i-Estadella, P., & Martín-Fernández, J. A. (2014b). Individual T² control chart for compositional data. Journal of Quality Technology, 46(2), 127–139. https://doi.org/10.1080/00224065.2014.11917958 google scholar
41. Wang, J., & Chen, Y. (2023). Introduction to transfer learning: Algorithms and practice. Springer Nature. google scholar
42. Yavuz, A., & Çilengiroğlu, Ö. V. (2020). Lojistik regresyon ve CART yöntemlerinin tahmin edici performanslarının yaşam memnuniyeti verileri için karşılaştırılması. Avrupa Bilim ve Teknoloji Dergisi, 18, 719–727. https://doi.org/10.31590/ejosat.691215 google scholar
43. Zaidi, F. S., Castagliola, P., Tran, K. P., & Khoo, M. B. C. (2019). Performance of the Hotelling T² control chart for compositional data in the presence of measurement errors. Journal of Applied Statistics, 46(14), 2583–2602. https://doi.org/10.1080/02664763.2019.1585690 google scholar
44. Zaidi, F. S., Castagliola, P., Tran, K. P., & Khoo, M. B. C. (2020). Performance of the MEWMA-CoDa control chart in the presence of measurement errors. Quality and Reliability Engineering International, 36(7), 2411–2440. https://doi.org/10.1002/qre.2705 google scholar