Big Data Application on Prediction of HDD Manufacturing Process Performance

Abstract

Every industry including manufacturing is going through digital transformation in this digital era we are in today. The digitization has transforming manual manufacturing to a highly connected end-to-end manufacturing leveraging on IoT and sensors. One of the common focuses in manufacturing is product yield and quality. However, lack of supplier’s data due to the manufacturing data generated is not shared and integrated into a common data platform becomes challenge among manufacturer to improve product’s yield and quality. The objective of this paper is to demonstrate the application of big data analytics on Hard Disk Drive (HDD) manufacturing and supplier data to predict HDD product yield/performance. In this research, the supplier data is focused on Head (read/write data) and Media (store data) component because it is critical to HDD’s performance in achieving the targeted storage capacity (i.e., Terabytes). The scope of this study includes data integration requirements and methods between supplier and HDD manufacturer, data platform to store and retrieve the integrated data, and the feasibility of predicting HDD product yield/performance using analytics with supplier’s data using Machine Learning. At the end of this research, target user groups will gain the ability to store and retrieve the integrated data between supplier and manufacturer from a big data platform, and to optimize manufacturing process for HDD performance using the insights and predictions generated by machine learning models.