TU Berlin

Sustainable Electric Networks and Sources of EnergyFUture Smart Energy (FUSE)


Page Content

to Navigation

FUture Smart Energy (FUSE)

Planned time frame:
August 2018 - December 2021

FUSE explores and demonstrates the application of AI technologies to increase the resilience and stability of the energy network with increasing percentage of renewables. The future grid will have a cellular structure with intelligent ICT-driven assets to balance local production and consumption, control intercellular energy flows and minimize system downtime. For this purpose, a hierarchical data flow architecture for distributed data processing is being developed.


The project explores and demonstrates the application of SoA AI technology to increase the resilience and stability of the future energy grid, which has to support the incorporation of an increasing percentage of renewable energy sources. This future grid will have a cellular structure where generators are closer to the consumers, with assets in the cells monitored and controlled by smart ICT. The main objective of FUSE is the development and demonstration of innovative ICT-based solutions for Demand Side Management, Condition Monitoring and Predictive Maintenance, which support the implementation of a cellular structure. FUSE achieves this goal by developing a hierarchical, highly scalable ICT architecture, AI-based distributed data management across all hierarchical levels, and decentralized intelligent control of consumers and network infrastructure.

Concept of the hierarchical FUSE ICT infrastructure for MV and LV grids

SENSE benefits the FUSE project with expertise and know-how in the areas of energy networks and demand side management focussing on modeling, simulation and optimization. The SENSE Smart Grid Laboratory serves as a development and validation environment as well as a demonstrator of the FUSE pilot implementation.

Cellular grid architecture

The focus of SENSE’s work in FUSE is the smart Demand Side Management: based on an initial catalog of requirements from the perspective of the low-voltage grid, specific models of flexible consumer loads are developed and implemented, which can be used for the smart DSM concept. In the second focus, TU Berlin supports the development of AI-based data management: Scenarios are created and extensive series of measurements are performed in the Smart Grid Laboratory, which provide the required training data for the machine learning algorithms developed by DFKI. After the training, the ML algorithms will be tested for effectiveness in the real infrastructure environment of the Smart Grid Laboratory.

As part of the project, the following functionalities are developed:

  • Development of a methodology for considering network-specific variables in the FUSE DSM concept
  • Development of functional models ("digital twins") for the flexible loads in the German pilot implementation, based on the DSM algorithms
  • Implementation and demonstration of the FUSE pilot in the SENSE Smart Grid Low Voltage Laboratory
  • Testing the developed algorithms of machine learning in FUSE pilots
  • Testing the developed DSM algorithms in conjunction with the medium-voltage pilot implementation in Finland

Project partners



Quick Access

Schnellnavigation zur Seite über Nummerneingabe