Augmented classification for electrical coil winding defects

• DOI: 10.1007/s00170-022-08671-w
• Publication Date: 2022-01-24T00:05:01Z
• Authors: Michael Farnsworth; Divya Tiwari; Ze Zhang; Geraint W. Jewell; Ashutosh Tiwari
• Keywords: Traceability

Abstract A green revolution has accelerated over the recent decades with a look to replace existing transportation power solutions through adoption of greener electrical alternatives. In parallel digitisation manufacturing enabled progress in tracking and traceability processes improvements fault detection classification. This paper explores machine manufacture challenges faced identifying failures modes during this life cycle demonstration state-of-the-art vision methods for classification coil winding defects. We demonstrate how generative adversarial networks can be used augment training these models further improve their accuracy challenging task. Our approach utilises pre-processing dimensionality reduction boost performance model from standard convolutional neural network (CNN) leading significant increase accuracy.