Digitalisation for Sustainability with Energy++

The goal of the transfer project between the University of Passau and Atos "Energy++" is to survey energy consumption at the University of Passau in order to determine indicators for strategic energy optimization.

The goal of the project "Energy++" is not only to collect, analyze and optimize the conventional energy consumption, but also the detailed (high resolution and data quantity) energy consumption of universities. The University of Passau provides itself as a model university. The transfer of KPIs known in the industry to the university context and the creation of awareness and best practices for sustainable IT at universities are strategic goals here. Another strategic goal is to exchange the collected data and expertise with Atos in order to promote sustainable universities and university datacenters.

In the first phase, a proof-of-concept is developed. The goal of the proof-of-concept is to implement a data mart or datawarehouse concept for the collection, analysis, and evaluation of energy data. Firstly, this enables the recording of the actual state of the University of Passau (on a small scale) in order to be able to plan further measures. Furthermore, the data quality and quantity can be assessed by Atos to see if they are sufficient for their own experience or if more data needs to be collected or the quality needs to be adjusted. This project phase was started in 2021, with two master theses "Analysis of energy efficiency in the workplace" being jointly supervised by the University of Passau (Professor Harald Kosch, Dr Armin Gerl) and Atos. The master thesis "Analysis of Energy Efficiency at the Workplace" deals with the use of different sensors to measure the power consumption of workplace IT. The goal is to build an experimental landscape with which various experiments can be conducted to save electricity, possibly by changing behavior. The master thesis "Sustainable AI in Data Centers" investigates energy and performance data of AI servers under load of different AI experiments, i.e. central IT infrastructure. The goal is to optimize the use of AI servers or to reduce energy consumption. A workflow for the collection, processing and genesis of KPIs and visualization is to be established inboth master theses and will serve as a basis for further work.

Symbolic picture: Adobe Stock