Search for Author, Title, Keyword
Milling cutter fault diagnosis using unsupervised learning on small data: A robust and autonomous framework
More details
Hide details
COEP Technological University, India
Institute of Fluid Flow Machinery, Polish Academy of Sciences,, Poland
Selcuk University, Turkey
These authors had equal contribution to this work
Submission date: 2023-09-19
Final revision date: 2023-11-24
Acceptance date: 2024-01-04
Online publication date: 2024-01-06
Publication date: 2024-01-06
Corresponding author
Rohan N Soman   

Institute of Fluid Flow Machinery, Polish Academy of Sciences,, 14, Fiszera Street, 80-231, Gdansk, Poland
Eksploatacja i Niezawodność – Maintenance and Reliability 2024;26(1):178274
  • Multi-class classification in 6 different classes pertaining to different faults commonly occurring in milling tool cutter.
  • Development of methodology based on unsupervised learning with very little training data needed.
  • Robustness of methodology shown with testing with blind data set from a different set of tool inserts.
  • Sensitivity studies to prove the robustness of the chosen parameters.
Tool condition affects the tolerances and the energy consumption and hence needs to be monitored. Artificial intelligence based data-driven techniques for tool condition determination are proposed. Unfortunately, the data-driven techniques are data-hungry. This paper proposes a methodology for classification based on unsupervised learning using limited unlabeled training data. The work presents a multi-class classification problem for the tool condition monitoring. The principal component analysis (PCA) is employed for dimensionality reduction and the principal components (PCs) are used as input for classification using k-means clustering. New collected data is then projected on the PC space, and classified using the clusters from the training. The methodology has been applied for classification of tool faults in 6 classes. The use of limited input parameters from the user makes the method ideal for monitoring a large number of machines with minimal intervention. Furthermore, due to the small amount of data needed for the training, the method has the potential to be transferable.
Dr Rohan Soman was supported by the project, ‘Guided waves based reference-free structural health monitoring using fiber Bragg grating sensors (REF-FREE)’ (2020/39/ST8/00188), granted by National Science Center, Poland.
Journals System - logo
Scroll to top