Employing Virtual Power Analytics and Linked Data for Enterprise IT Energy Informatics

The paper provides a solution to an organizations challenge for reducing its ecological footprint. We develop a frame- work for visualizing information concerning power consump- tion of IT devices in real-time using power metering ap- proaches and leverage semantic web technologies to weave information that is dispersed across different data sources, ranging from an organizations asset databases to excel spread- sheets to the data available on manufacturers websites. Vi- sualizing the aforementioned information and bringing it together in one application can help organizations in understanding and adapting their energy consumption behavior in order to better support environmentally sustainable practices. The purpose of this project, from a power analytics standpoint, is twofold: (1) it aims to provide users with insight into their devices current power consumption without leaving a footprint on their machine, and (2) to analyze if one method gives more accurate results than others. We performed analysis using Microsoft Joulemeter and power estimation using software design methodologies (with numbers from smart meters as our benchmark).