Research Unit G: Trend research for long-lived and efficient hydrogen fuel cell systems

The DuraFuelCell joint project, funded by the DFG, aims to develop and better understand long-lived, efficient, hydrogen fuel cell systems for mobile and stationary applications. The Data-based trend and scenario research subproject makes an important contribution to this goal by developing an AI-based knowledge database that serves as a central platform for structuring international research activities, promoting knowledge exchange, and optimising research activities in the field of hydrogen fuel cell technology.

Hydrogen is currently experiencing an upswing as a deciding component in the decarbonisation of our energy supply. Both the German national and the Bavarian governments are planning the transition of important CO2-intensive industries, including the transport sector, to a basis of green hydrogen. Sustainable, hydrogen-based fuel cells are essential for the reconversion of hydrogen into electrical energy; and the long-life cycle of these systems will be decisive. Considering the rapid progress in this field, numerous approaches to this problem will be considered. However, it can be challenging to determine which approaches actually will lead to sustainable technological and economic development. Progress is often published in a variety of channels and with high frequency, which makes it challenging for research institutions, but also for companies and especially small and medium enterprises (SMEs) to identify and track relevant developments and subsequently to integrate them into their own innovation processes.

The aim of this subproject is to make the complexity of the developments in the field of hydrogen-based fuel cells manageable through an AI-based approach to dynamically record and structure international research activities and initiatives. This continuing survey of current research in the fuel cell field enables the identification, classification, and monitoring of relevant developments, technologies, and actors. The structured representation of this knowledge, with additional links to existing public knowledge databases, will facilitate the identification of trends and dependencies in technological developments. This will provide usable findings for specific questions in hydrogen fuel cell research. 

To accomplish this goal, we will design and develop a methodology for automated knowledge extraction, representation, and curation. This will primarily comprise:

  • The development of a domain-specific ontology for the fuel cell field in collaboration with domain experts. This ontology will facilitate the classification of actors and technological developments and to the identification of relationships or dependencies within those groups. 
  • The application of NLP-based algorithms for the automatic extraction of structured data from text sources.  
  • The integration of information from differing sources (e.g., extracted data from texts, public databases, expert knowledge) in a semantically enriched, dynamic knowledge graph
  • The facilitation of visualisation and interaction with the knowledge graph through a user interface

The knowledge database developed within this subproject will grow, keep specialist knowledge easily accessible, and generate synergies in the individual research areas.  This transparency will support the evaluation of the research results and the identification of further research areas.

 

More information about Prof. Blum’s group (in German).

Principal investigator

Name Contact
Ralph Blum Ralph Blum
Prof. Dr.

Research associate