Detection of tool wear in drilling using acoustic machine learning models through labeling-threshold sensitivity analysis
DOI:
https://doi.org/10.24054/rcta.v2i48.4535Keywords:
tool wear, CNC drilling, ordinal classification, acoustic analysis, SVM, threshold sensitivity, predictive maintenance, Industry 4.0Abstract
An ordinal classification system for drill bit wear in CNC drilling of AISI 4140 steel using acoustic signals captured with low-cost microphones is presented. Frank-Hall ordinal decomposition is employed with SVM and RF over 10 statistically selected acoustic features. The central contribution is a labeling threshold sensitivity analysis between 60% and 97% of tool life. Adjacent accuracy stays above 83% across the entire range, meaning errors never exceed one ordinal step. A refined SVM (linear kernel, C=5, asymmetric weights, calibrated thresholds) reaches adjacent accuracy 0.985 and exact accuracy 0.602 at the 75% threshold, surpassing the baseline (0.859 and 0.503). Spectral gating reduces the cross-microphone domain shift (F1: 0.18 to 0.33). Data, feature-extraction scripts and trained models are openly available in the repository associated with this work.
Downloads
References
O. I. Ayala Ortiz, “Base de conocimiento para detección de desgaste de herramienta,” 2026, Zenodo/GitHub. Accessed: Jun. 08, 2026. [Online]. Available: https://github.com/AyalaOrtizOscar/Base-de-conocimiento/tree/main
M. P. . Groover, Fundamentos de manufactura moderna: materiales, procesos y sistemas. McGraw Hill, 2007.
M. S. H. Bhuiyan, I. A. Choudhury, and M. Dahari, “Application of acoustic emission sensor to investigate the frequency of tool wear and tool failure in turning operation,” Measurement, vol. 92, pp. 208–217, 2016, doi: 10.1016/j.measurement.2016.06.006.
M. C. Shaw, Metal Cutting Principles, 2nd ed. Oxford University Press, 2005.
Y. Restrepo Chaustre, J. A. Becerra Vargas, and A. Pardo García, “Metodología de detección de fallas de un motor síncrono methodology for fault detection of a synchronous motor,” Revista Colombiana de Tecnologías de Avanzada, 2017.
J. E. Meneses Flórez and N. Peña, “Estudio correlacional entre la señal acústica y el desgaste de la herramienta en un proceso de taladrado,” Revista Colombiana de Tecnologías de Avanzada (RCTA), vol. 2, no. 34, pp. 1–8, 2020.
N. Orejarena Osorio and N. Peña García, “Estudio de la señal acústica en función de la vida de la herramienta en un proceso de taladrado,” Trabajo de grado, Universidad Industrial de Santander, Bucaramanga, Colombia, 2014.
I. E. Pacheco Córdova, R.-V. Sánchez Loja, D. Cabrera, and M. Cerrada, “Adquisición de señales de vibración y emisión acústica para el diagnóstico de severidad de fallos en maquinaria rotativa,” Revista Colombiana de Tecnologías de Avanzada (RCTA), vol. 3, no. 1, pp. 87–94, Oct. 2018, doi: 10.24054/rcta.vi.156.
Y. LeCun, Y. Bengio, and G. Hinton, “Deep learning,” Nature, vol. 521, no. 7553, pp. 436–444, 2015, doi: 10.1038/nature14539.
L. Wang, M. Törngren, and M. Onori, “Current status and advancement of cyber-physical systems in manufacturing,” J. Manuf. Syst., vol. 37, pp. 517–527, 2015, doi: 10.1016/j.jmsy.2015.04.008.
D. A. Dornfeld and D.-E. Lee, Precision Manufacturing. New York: Springer, 2008.
E. Frank and M. Hall, “A simple approach to ordinal classification,” in Proceedings of the European Conference on Machine Learning (ECML), Springer, 2001, pp. 145–156.
O. Iván Ayala Ortiz, “Base de conocimiento para monitoreo acústico,” Universidad Industrial de Santander, 2026.
V. MacPhail, J. Buxton, F. Bussière, and A. C. Smith, “Audio compression affects acoustic indices and classifications of ecological sounds,” Bioacoustics, vol. 33, no. 1, pp. 1–22, 2024.
T. Sainburg, M. Thielk, and T. Q. Gentner, “Finding, visualizing, and quantifying latent structure across diverse animal vocal repertoires,” PLoS Comput. Biol., vol. 16, no. 10, Art. no. e1008228, 2020, doi: 10.1371/journal.pcbi.1008228.
Y. Li, J. Li, Y. Zhang, and Z. Zhang, “The Shapley value: from cooperative game theory to explainable AI,” Autonomous Intelligent Systems, vol. 4, Art. no. 2, 2024.
T. Banda, A. Ali, J. Asku, and H.-C. Möhring, “Multi-sectional SVD-based machine learning for tool wear assessment and prediction in turning process,” International Journal of Advanced Manufacturing Technology, vol. 132, no. 7–8, pp. 3585–3600, 2024, doi: 10.1007/s00170-024-13610-y.
Downloads
Published
Issue
Section
License
Copyright (c) 2026 Jorge Enrique Meneses Flórez, Nicolás Orejarena Osorio, Oscar Ivan Ayala Ortiz
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.
