Resource-efficient technologies for sustainable, global value creation

Powerful computers have become a fact of everyday life and the digital transformation is continually engendering new high-tech business models. While fulfilling our demand for quality of life, security and health, these advances consume a lot of energy. It is our conviction that forward-looking technologies for our economy and day-to-day lives must be in harmony with the needs of a sustainable world, which is why we see it as our mission to promote resource efficiency.

Microelectronics can help bring about notable energy savings 

Albert Heuberger: We are continuously developing and improving the performance of technologies and entire systems. While doing so, we strive to make the energy consumption of electronics as efficient as possible – through means such as energy-efficient AI processors. When it comes to microelectronics, we start with things like resource efficiency in sensor edge cloud systems, energy-saving communication infrastructures and resource-optimized electronics production.

Resource-efficient electronics is the basis for a forward-looking economy

Industry 4.0, machine learning and artificial intelligence are making new demands in relation to energy consumption, data processing and data transfer times. For this reason, we are working to advance next-generation computing and developing quantum-based and neuromorphic computing as well as resource-efficient microelectronic components and systems.

Green ICT @ FMD is a central contact point for green microelectronics 

On August 1, 2022, we founded Green ICT @ FMD, a cross-location competence center for resource-conscious information and communication technology. The center has been awarded funding of 34 million euros by the German Federal Ministry of Education and Research (BMBF) as part of the Green ICT initiative within the German government’s Cli-mate Action Program 2030. Green ICT hubs will be opened with a focus on sensor edge cloud systems, infrastructures, materials and processes for green production. The Research Fab Microelectronics Germany (FMD) will function as a central point of contact for a wide variety of environmental questions in electronics – for industry, for politics, for individual customers, for young people and for students.

Efficiently implemented algorithms give competitive advantage 

Bernhard Grill: For businesses to maximize their competitive advantage, they need to go beyond economical electronics. Resource-efficient technologies also require adapted, efficiently implemented algorithms for the respective platform. As such, companies need to carefully choose the right techniques and continuously need to exploit relevant processor-specific optimizations.

SPEAKER voice assistance system uses energy-efficient AI processors 

In the domain of voice assistance systems in particular, the SPEAKER project has made significant progress in the use of energy-efficient AI processors. The project is funded by the German Federal Ministry for Economic Affairs and Climate Action as part of its AI Innovation Competition.

AI-based algorithms efficiently process large amounts of information in real time 

The implementation of efficient algorithms and the increasing performance capabilities of microelectronics are allowing us to develop new technologies and processes that also technically improve our products. A good example of this is Fraunhofer upHear Spatial Audio Microphone Processing. It provides spatial audio capturing with innovative and automatic noise suppression that eliminates interfering background noises: naturally, efficiently and in real time for very compact devices.

Moreover, advanced AI techniques are being used to synthesize natural sounding language from written text. These applications must efficiently process large amounts of information in real time. Fraunhofer IIS has made significant advances in this area with its Allinga Voice speech synthesis solution.

Advancing Industry 4.0 through resource efficiency 

Alexander Martin: In addition to resource-efficient hardware design, the resource efficiency of the algorithms that run on this hardware plays an important role. Algorithms, based on AI and mathematical optimization models, are a vital part of countless applications in Industry 4.0 today. Their reliability critically depends on the data quality, but also on the volume of data. The generation and storage of this data consumes huge amounts of energy. Handling data in a resource-conserving, sustainable manner in each phase of the data life cycle is therefore essential for the digitalization of tomorrow. 

Optimization algorithms permit energy savings 

Beyond the energy-efficient design of the algorithms themselves, the solutions calculated using the algorithms can also enhance energy efficiency in a wide variety of industrial applications. Digitalization can therefore be an opportunity to make a significant contribution to energy savings and thus to the reduction of CO2 emissions in many domains by means of the intelligent control and management of systems, processes and networks.

The goal here is not just enhancing the control of established energy networks and production processes, but also further developing energy systems and determining the optimal energy mix for the energy supply of the future.