EnergySHR matches supply and demand of data and algorithms for the energy transition.
Access to representative data is critical for developing new algorithms to help plan, design and manage the energy system of the future.
Delft University of Technology, Erasmus University Rotterdam and Erasmus University Medical Center have joined forces in Convergence. The urgent and complex societal challenges of our time require fundamentally new forms of collaboration, in which collaboration takes place across the boundaries of institutes and disciplines to develop new perspectives and solutions. Convergence is built around five programmes with each their own focus, one of which is Artificial Intelligence (AI).
At the initiative of the AI Convergence Center for Energy System Intelligence, the EnergySHR platform has been created for matching the supply and demand between data and algorithms. It means to tackle the various challenges related to privacy and confidentiality that stakeholders and data owners in this domain experience. EnergySHR wishes to lower the barriers for collaboration and accelerate data-driven innovation.
What is the EnergySHR platform?
The EnergySHR platform is at the level of ‘minimal viable product’, to explore together with the user group what important features and use cases are for further development. With this input, the platform is further developed and the user base is gradually expanded.
The goal is to allow data scientists, data consumers and data providers (like grid operators, local government, energy service providers and energy companies) to share, collaborate and commercialize their data services. EnergySHR facilitates this in a secure and privacy-friendly manner and ensures that data owners stay in full control over their data.
EnergySHR is based on the Gaia-X concepts which is part of the Digital Europe Programme aiming at making Europe’s digital infrastructure future-proof and self-sovereign.
- Digital Sovereignty – Data owner stays in control
- Preventing copies of data – Data stays at the source
- Labels – Digital quality stamps for datasets and algorithms
- Sharing options for both open and restricted data
Artificial intelligence to generate synthetic data
With BlueGen.ai’s software, EnergySHR users are able to generate synthetic versions of data sets that will be no longer traceable to confidential and personal information from individual records. The software runs on Artificial Intelligence that learns from the original data and creates synthetic data sets that look and behave the same as the underlying real data. They are suitable for experimental research and machine learning development, without the risk of leaking sensitive information. By giving the option to publish synthetic data instead of the real data, EnergySHR lowers the barriers for data owners to fill the platform with data and create traction in the early stages of the platform
Publishing synthetic data on EnergySHR
Data owners are in full control of their data sets regarding accessibility and usage. They may choose to publish either real data, synthetic data or both. A published synthetic data set is always accompanied by an evaluation of the resemblance, utility and privacy of the data and instructions on how to use it with a disclaimer.
Synthetic data in practice
The option of synthetic data on the EnergySHR platform opens up opportunities for both the supply (data owner) and demand (data user) side, for example:
- As an energy company, you want to share the statistical value of the data set, but not the individual records due to confidentiality or privacy restrictions
- As a researcher, you want to research the statistics in a data set without a very specific scope
- As a data scientist, you want to experiment freely with your own models and software on the data
“BlueGen.ai supports EnergySHR in its goals to accelerate data-driven innovation for the energy transition, both with the practical appliance of AI and the resulting usefulness of synthetic data”
Caroline Duterloo,
Innovation Manager Artificial Intelligence – TU Delft
