The adoption of the Internet of Things (IoT), smart environments, and Industry 4.0 requires highly distributed data processing close to the edge of the network, known as edge or fog computing. By enabling local data processing through the geographic proximity of computing nodes and sensors or end devices, we can reduce latency, save bandwidth, and ensure data confidentiality. An important application domain of edge computing is artificial intelligence, also known as edge intelligence. Novel machine learning approaches, such as federated learning, ensure privacy in edge scenarios but result in additional energy usage at often energy-constrained hardware.
A similar paradigm shift as from centralized cloud to highly distributed edge computing is taking place in the energy industry. Whereas previously most energy was provided by large, centralized producers such as nuclear and coal-fired power plants, energy generation is becoming increasingly decentralized as part of the transition to renewable energy sources. Solar and wind power plants are spreading over large areas and penetrating urban spaces. However, managing energy systems is becoming increasingly difficult as distribution and variability increase. Energy storage systems can only partially mitigate this effect, so research and industry are focusing on adapting the demand side to green energy availability as well.
The SYNERGY project is a collaboration of DOS and Huawei funded by Software Campus. We seek to identify and exploit synergies in the areas of edge intelligence and distributed power generation by means of co-simulation and experimentation on real hardware. The overall goal is to reduce the environmental footprint of future edge intelligence applications and contribute to CO2-neutral AI.