Algorithmic assessment of drag on thermally cut sheet metal edges
-
Janek Stahl
,
Simon Zengl
Abstract
Drag is a key criterion in assessing the quality of thermally cut sheet metal edges, which is critical to the reliability of the final product. The evaluation of drag has been described qualitatively and quantitatively, but the scientific literature lacks a methodical description of algorithmic tracking of the drag lines themselves. This absence of a standardized approach has hindered the objective determination of drag. With recent advances in the field towards automated quality assessment aimed at autonomous adaptation of process parameters, the need for consistent, fast and reliable assessment of drag lines has become apparent. To address this gap, this study introduces an innovative drag line tracking algorithm, inspired by the behavior of fluid flowing towards the lowest points, to compute a generalized drag line for an edge with a homogeneous cutting pattern. The algorithm utilizes the height data of the measured cut edges as a data base for the assessment of the drag lines. The results indicate that the drag lines identified by the algorithm are not only subjectively accurate, but also show a strong correlation with human-annotated drag lines across several metrics. This work lays the foundation for the objective evaluation of drag by not only describing an algorithm for the consistent determination of drag lines, but also by presenting a tool for human annotation and suitable customized metrics. As a result, it contributes significantly to the comprehensive evaluation of edge quality and represents a step forward in the automatic optimization of process parameters and the improvement of cutting edge quality.
Zusammenfasung
Der Rillennachlauf ist ein entscheidendes Kriterium bei der Beurteilung der Schnittkantenqualität von thermisch geschnittenen Blechen, die für die Zuverlässigkeit des Endproduktes von entscheidender Bedeutung ist. Die Bewertung des Rillennachlaufs ist qualitativ und quantitativ beschrieben worden, jedoch fehlt in der wissenschaftlichen Literatur eine methodische Beschreibung der algorithmischen Verfolgung der Rillen selbst. Das Fehlen eines standardisierten Ansatzes hat die objektive Bestimmung des Rillennachlaufs erschwert. Mit den jüngsten Fortschritten auf dem Gebiet der automatisierten Qualitätsbewertung, die auf eine autonome Anpassung der Prozessparameter abzielt, ist der Bedarf an einer konsistenten, schnellen und zuverlässigen Bewertung des Rillenverlaufs deutlich geworden. Um diese Lücke zu schließen, wird in dieser Studie ein innovativer Algorithmus zur Verfolgung von Rillen eingeführt, der auf dem Verhalten von Flüssigkeiten basiert, die entlang der tiefsten Punkte fließen, um einen verallgemeinerten Verlauf von Rillen für eine Kante mit einem homogenen Schnittmuster zu berechnen. Der Algorithmus verwendet die Höhendaten der gemessenen Schnittkanten als Datenbasis für die Bestimmung des Rillenverlaufs. Die Ergebnisse zeigen, dass die mit dem Algorithmus berechneten Rillenverläufe nicht nur subjektiv genau sind, sondern auch in mehreren Bereichen eine hohe Korrelation zu den von Menschen markierten Rillenverläufen aufweisen. Diese Arbeit legt den Grundstein für eine objektive Bewertung des Rillennachlaufs, indem sie nicht nur einen Algorithmus zur konsistenten Bestimmung des Rillennachlaufs beschreibt, sondern auch ein Werkzeug zur menschlichen Annotation und geeignete maßgeschneiderte Metriken vorstellt. Damit leistet sie einen wesentlichen Beitrag zur ganzheitlichen Bewertung der Kantenqualität und stellt einen Schritt zur automatischen Optimierung von Prozessparametern und zur Verbesserung der Schnittkantenqualität dar.
About the authors
M. Sc. Janek Stahl received his bachelor’s degree in mechanical engineering from HTWG Konstanz in 2009 and his master’s degree in mechanical engineering from the University of Stuttgart in 2014. Since then, he has been conducting research at the Fraunhofer Institute for Manufacturing Engineering and Automation IPA in Stuttgart, Germany. In the Artificial Intelligence and Machine Vision department, he focuses on machine learning and texture analysis in the field of 2D image processing for industrial applications.
Simon Zengl is studying mathematics at the University of Stuttgart since 2018. He is a student research assistant in the Department Machine Vision and Signal Processing at the Fraunhofer Institute for Manufacturing Engineering and Automation IPA in Stuttgart since 2020, where he is mainly involved in machine learning for image processing.
Dr.-Ing. Andreas Frommknecht received his Diploma in Mathematics from University of Ulm in 2011. Since 2011, he has started working as a research associate in the Department Machine Vision and Signal Processing at the Fraunhofer Institute for Manufacturing Engineering and Automation IPA. Since 2020, he leads the group Optical Measurement and Testing Systems in the same department. His research focus is on analytical and machine learning based visual data processing.
M. Sc. Christian Jauch completed his degree in technical cybernetics at the University of Stuttgart in 2015. Since then, he has been a researcher at the Fraunhofer Institute for Manufacturing Engineering and Automation IPA. Since 2021, he leads the Activity and Scene Analysis research group focusing on artificial intelligence and machine vision. His work emphasizes human-centric process optimization. Key applications include assembly assistance, employee training, automated documentation, and contactless control, aiming to improve efficiency and ease in industrial processes.
Prof. Dr.-Ing. Marco F. Huber received his diploma, Ph.D., and habilitation in computer science from Karlsruhe Institute of Technology (KIT). He led a research group at Fraunhofer IOSB and was Senior Researcher at AGT International until 2015. He then led product development at USU Software AG and was adjunct professor at KIT. Since 2018, he is full professor at the University of Stuttgart and scientific director at Fraunhofer IPA. His research focuses on machine learning, image processing, and robotics.
-
Research ethics: Not applicable.
-
Informed consent: Not applicable.
-
Author contributions: The authors have accepted responsibility for the entire content of this manuscript and approved its submission. J.S. developed the algorithm and wrote most of the paper. S.Z. developed the program for human annotation and calculated the comparisons with the corresponding metrics. A.F. reviewed the manuscript. C.J. reviewed the manuscript. M.H. reviewed the manuscript.
-
Use of Large Language Models, AI and Machine Learning Tools: LLMs were utilized to assist in formatting graphics and tables in LaTeX, as well as to suggest improvements for the accuracy of the text. Additionally, Deepl, an AI-based translation tool, was employed to refine wording and phrasing for clarity.
-
Competing interests: The authors state no competing interests.
-
Research funding: None declared.
-
Data availability: The raw data can be obtained on request from the corresponding author.
References
[1] L. Tatzel and F. Puente León, “Impact of the thermally induced focus shift on the quality of a laser cutting edge,” J. Laser Appl., vol. 32, no. 2, 2020. https://doi.org/10.2351/7.0000053.Search in Google Scholar
[2] T. Himmer, T. Pinder, L. Morgenthal, and E. Beyer, “High brightness lasers in cutting applications,” Int. Congr. Appl. Lasers Electro-Opt., vol. 2007, no. 1, p. 206, 2007. https://doi.org/10.2351/1.5061055.Search in Google Scholar
[3] J. Stahl and C. Jauch, “Quick roughness evaluation of cut edges using a convolutional neural network,” in Fourteenth International Conference on Quality Control by Artificial Vision, vol. 11172, 2019, pp. 175–181.10.1117/12.2519440Search in Google Scholar
[4] A. Krizhevsky, I. Sutskever, and G. E. Hinton, “Imagenet classification with deep convolutional neural networks,” Adv. Neural Inf. Process. Syst., vol. 25, 2012.Search in Google Scholar
[5] L. Tatzel and F. Puente León, “Image-based roughness estimation of laser cut edges with a convolutional neural network,” Proc. CIRP, vol. 94, pp. 469–473, 2020. https://doi.org/10.1016/j.procir.2020.09.166.Search in Google Scholar
[6] N. Levichev, A. T. García, M. Kardan, D. Cattrysse, and J. R. Duflou, “Experimental validation of a machine learning algorithm for roughness quantification in laser cutting,” Proc. CIRP, vol. 113, pp. 564–569, 2022. https://doi.org/10.1016/j.procir.2022.09.175.Search in Google Scholar
[7] L. Tatzel, O. Al Tamimi, T. Haueise, and F. Puente León, “Image-based modelling and visualisation of the relationship between laser-cut edge and process parameters,” Opt Laser. Technol., vol. 141, 2021, Art. no. 107028. https://doi.org/10.1016/j.optlastec.2021.107028.Search in Google Scholar
[8] J. Stahl, C. Jauch, M. Tuncel, and M. F. Huber, “Investigation of different illumination scenarios for the evaluation of thermally cut edges with convolutional neural networks using a mobile device,” Electron. Imaging, vol. 33, pp. 1–7, 2021. https://doi.org/10.2352/issn.2470-1173.2021.6.iriacv-314.Search in Google Scholar
[9] J. Stahl, A. Frommknecht, and M. F. Huber, “Comprehensive quantitative quality assessment of thermal cut sheet edges using convolutional neural networks,” in British Machine Vision Conference 2023, 2023.Search in Google Scholar
[10] International Standard, ISO 9013 Third Edition 2017-02: Thermal Cutting — Classification of Thermal Cuts — Geometrical Product Specification and Quality Tolerances, 2017.Search in Google Scholar
[11] L. Budach, et al.., “The Effects of Data Quality on ML-Model Performance,” CoRR abs/2207.14529, 2022.Search in Google Scholar
[12] D. C. Montgomery, Design and Analysis of Experiments, Hoboken, NJ, Wiley, 2017.Search in Google Scholar
[13] P. E. Shrout and J. L. Fleiss, “Intraclass correlations: uses in assessing rater reliability,” Psychol. Bull., vol. 86, no. 2, pp. 420–428, 1979. https://doi.org/10.1037//0033-2909.86.2.420.Search in Google Scholar PubMed
[14] P. Newbold, W. L. Carlson, and B. M. Thorne, Statistics for Business and Economics, Harlow, UK, Pearson, 2013.Search in Google Scholar
[15] I. Lawrence and K. Lin, “A concordance correlation coefficient to evaluate reproducibility,” Biometrics, vol. 45, no. 1, pp. 255–268, 1989. https://doi.org/10.2307/2532051.Search in Google Scholar
© 2024 Walter de Gruyter GmbH, Berlin/Boston
Articles in the same Issue
- Frontmatter
- Research Articles
- Measurement and advanced data post-processing of proton resonance frequency shift in 7 T MRI to obtain local temperature in a tissue-mimicking phantom
- Algorithmic assessment of drag on thermally cut sheet metal edges
- Messunsicherheit geometrischer Prüfmerkmale – Automatisiert und praxisgerecht mit Koordinatenmessgeräten ermitteln
- Measuring the frequency response of an optical microphone system with a fiber based setup
- Spatial detection and localisation of multiple laser beams in optical measuring systems
Articles in the same Issue
- Frontmatter
- Research Articles
- Measurement and advanced data post-processing of proton resonance frequency shift in 7 T MRI to obtain local temperature in a tissue-mimicking phantom
- Algorithmic assessment of drag on thermally cut sheet metal edges
- Messunsicherheit geometrischer Prüfmerkmale – Automatisiert und praxisgerecht mit Koordinatenmessgeräten ermitteln
- Measuring the frequency response of an optical microphone system with a fiber based setup
- Spatial detection and localisation of multiple laser beams in optical measuring systems
