Abstract—When Materials are machined with subtractive manufacturing the material will be chipped resulting in scrapings of the base material. Those scrapings can get tangled up with the tool, leading to damages of the part and-or the tool. Currently, the machining will continue until the assigned worker notices the chip buildup. Then there already could be scrapings and damaged parts must be machined again. To prevent this issue, a mounted camera capture images of the machine tool. These pictures will be classified by an Image Classification model differentiating between tangled up and harmless scraps. A Residual Convolutional Neural Network was trained to learn the difference between the two classes on a very limited labeled dataset. When a harmful type of chip is recognized by the Image Classification model, preprogrammed measures can be initiated e.g., stopping the process, changing the tool, or changing the process parameters. This should contribute to the possibility to get the maximum service life from each tool without the need of highly specialized personal supervision of the whole machining process while also minimizing the chance of tool breakage.
Index Terms—Image classification, convolutional neural network, residual networks, data augmentation.
Joshua Hermann, Roman Radtke, and Alexander Jesser are with University of Applied Sciences Heilbronn, Germany (e-mail: joherman@stud.hs-heilbronn.de, roman.radtke@hs-heilbronn.de, alexander.jesser@hs-heilbronn.de, alexander.jesser@hs-heilbronn.de).
Aibek Kabanbayev and Junisbekov Mukhtar Shardarbekovich are with t Dulaty University Taraz, Kazakhstan, (e-mail: aibek.kabanbayev@hs-heilbronn.de, d_muhtar@mail.ru).
Cite: Joshua Hermann, Roman Radtke, Aibek Kabanbayev, Junisbekov Mukhtar Shardarbekovich, and Alexander Jesser, "KI Based Chip Classification for Detection of Unwanted Chip Buildup," International Journal of Future Computer and Communication vol. 11, no. 4, pp. 79-83, 2022.
Copyright © 2022 by the authors. This is an open access article distributed under the Creative Commons Attribution License which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited (CC BY 4.0).
Copyright © 2008-2025. International Journal of Future Computer and Communication. All rights reserved.
E-mail: editor@ijfcc.org