Abnormality Detection in Hard Disk Drive Assembly Process Using Support Vector Machine

Masayuti Simongyi and Prabhas Chongstitvatana

submit to ECTI-CON 2018, July 18 – 21, Chiang Rai, Thailand

Abstract

The research proposes a method to detect abnormality in assembly of a hard disk drive. Machine learning techniques are employed to recognise the behavior of good and bad components. Both good and bad components of disk drive assembly are collected from the assembly line. To test the components, the current in voice coil motor is measured and collected for using as training data set. Since the amount of abnormal drive in hard disk drive assembly process is very small, this paper also set the experiment of varying amount of training data set that can satisfy practical hard disk drive assembly process. Support Vector machine is chosen as it is very good in binary classification. It is found to be suitable for this task.