Innovation competitions: sustainable electronics and energy-efficient AI

The increasing use of electronics and artificial intelligence (AI) systems is helping to make considerable energy and resource savings.

However, these systems themselves also consume significant amounts of energy. The German Federal Ministry of Education and Research (BMBF) has launched two research competitions to find the smartest energy consumption solutions for the development of AI systems and for electronics for information and communication technologies, respectively. The winners will be awarded project funding to help make their ideas reality. 

Innovation competition “Electronics for saving energy in information and communication technology”

October 2020 – June 2021

GreenICT.jpg

© Adobe Stock/Oleg

Digital products and services play a significant role in saving energy and resources. However, smart solutions are needed to reduce the energy consumption of the products and services themselves. The BMBF’s Green ICT initiative is part of the Climate Action Programme 2030. The initiative aims to fund the research and development of electronic components and systems that will help save energy and carbon dioxide in information and communication technologies (ICT).

As part of the Green ICT Initiative, the innovation competition “Electronics for saving energy in information and communication technology” was launched on 1 October 2020. Over the next nine months, ten teams of research organisations and universities will compete against one another to develop the solution that will have the greatest impact on making the digital transformation more sustainable. The jury will judge the projects based on the following criteria: potential for innovation in electronics, prospect of widespread industrial application and, of course, potential to reduce carbon dioxide emissions. The winners will receive a further three years of funding from the BMBF to support the realisation of their ideas in collaboration with industrial partners.

The overall aim is to encourage a rapid transfer of ideas about sustainable digitalisation from research into widespread use. The competition also aims to help expedite the implementation of the Climate Action Programme 2030.

For further information please visit the German-language web page describing the GreenICT innovation competition.

Pilot innovation competition “Energy-efficient AI system”

October 2019 – December 2020

bkm energieeffiziented ki system 2019
© Adobe Stock / trahko

AI is becoming more and more common in all aspects of our lives. Often, it is clearly labelled “AI”, but it is also frequently hidden within complex applications. Over the last few years, AI has become a key technology for innovations in almost every possible field of application. The Federal Government pursues a holistic approach of AI within its Artificial Intelligence Strategy and intends to take an active role in shaping the future development of this technology.
One of the downsides of the current AI boom is the significant increase in the amount of energy consumed by information and communication technologies. A common approach to improving the energy efficiency of AI hardware is to use specialized processors or accelerators in large-scale data centres. However, while transferring data to a data centre makes a significant difference in increasing energy efficiency, another solution might be to process AI on mobile devices. AI calculations are often run on mobile devices close to the user for reasons of data security, for time-critical applications and to avoid problems caused by interrupted data connections. However, at the moment, mobile AI hardware is still too energy-intensive.

In conjunction with the Federal Agency for Disruptive Innovation (SPRIN-D), the BMBF aims to address this issue with the “Energy-efficient AI system” competition. Eleven teams from universities and research institutions are currently competing against one another to build the most energy-efficient system that will use their disruptive concepts for AI-specific accelerators to solve a medical problem. The systems must be able to identify cardiac arrhythmia in real electrocardiogram data with guaranteed precision. The competition has three categories based on different basic technologies. Each of the winners will receive funding for a subsequent research project to help further develop their ideas. The challenge shines a spotlight on Germany’s academic expertise in the field of AI hardware and aims to close the gap between academia and the AI user industry, which is also represented on the competition’s board of advisors.

The contest is set to run until December 2020. The winners will be chosen by assessing and comparing the systems developed by the participants. The award ceremony is scheduled for March 2021.

For further information please visit the German-language web page describing the Pilot innovation competition “Energy-efficient AI system”.