Click on each author name to view the details.
Failure of a data center's UPS system can mean substantial losses, and batteries are consistently a leading root cause of those failures. The key to avoiding these losses is being able to accurately identify and service battery strings that are at risk - insights that can be extrapolated from data. This paper will focus on data-driven analysis by detailing one project in which a team used millions of data points, gathered over 12 years, to compare the operational performance of customers’ individual units to the historic performance of the same model in similar environments and applications. The paper will expound on the data acquisition and advanced big-data analysis process that enabled the team to accurately forecast remaining service life, including correlation of the effects of nine key external factors from the historic battery service life portfolio.