Data Pre-processing Methods of Hard Disk Drive Data for Industrial Data Center
Authors: Xu, L., De Vrieze, P. and Balogun, K.
Abstract:With the convergence of many technologies in Industry 4.0, more and better data are being captured, exchanged, and analysed. This large quantity of data creates opportunities that go beyond monitoring and improving efficiency, promoting significant changes in manufacturing, organizations, and society. Current trends towards Industry 5.0 introduce a sustainable, human-centric, and resilient industry. For supporting such a trend, handling a large quantity of data for supporting industrial applications requires a data center. In this paper, we reviewed data pre-processing methods for hard disk drive data (HDD) analysis for predictive maintenance (PdM) in a data center. The Backblaze HDD data is used to discuss the characteristics of HDD data. The notable analytical techniques are explored. We generate a framework for analyzing hard disk data for predictive maintenance. The framework guides how to select pre-processing methods for further data analysis models with different metrics.
Source: Manual