TEF EV Node

ev
Accelerating Smart EV Integration into Energy Communities
About
TEF EV is a dynamic testing environment focused on supporting the integration of electric vehicles (EVs) into diverse energy ecosystems, particularly Energy Communities (ECOs). It offers a real-life setting for testing and experimentation, enabling the development and adoption of AI-driven smart charging solutions, designed to adapt to user habits and ECO dynamics.
Testing Infrastructure
The testing site is based in Luxembourg and is provided by EnergiPark, whereas the computational resources are provided by LIST. On the virtual side, EMOT is responsible for developing AI services, EnergiPark is an end user that will enhance predictive and optimisation models, and lastly, AUTEL will serve as the main end-user and provide SEVCs that will be used to test the AI services.
Physical

EnergiPark is responsible for the assets of the ECOs, while LIST provides its testing facilities and HPC infrastructure.

Luxembourg

  • A power laboratory for testing SEVCs under various operational conditions
  • Α programmable dSPACE SCALEXIO smart EV charging control system
  • Power hardware-in-the-loop (PHiL) environment with:
    • 3 interconnected OPAL 5600 platforms (eMEGASIM, ePHASORSIM and eFPGASIM)
    • 2 interconnected OPAL 4610X systems with HYPERSIM
  • The AIDA platform, consisting of 34 servers, that will be expanded through Meluxina supercomputers that consists of additional 573 nodes
  • An accelerator module with 200 GPU nodes
  • An accelerator module with 20 FPGA nodes
  • 3 real sites with ECOs and districts with PV generation, manageable loads, and several community-oriented BESS
Virtual

Regarding the virtual infrastructure, EMOT brings its expertise in services based on the latest AI/ML techniques, while LIST provides a powerful cloud-based platform that will act as a central hub for digital components.

  • The cloud based INDIGENER® Digital Platform for hosting digital components
  • A Digital Twin of real-life testing sites
  • SEVC AI management models for descriptive, diagnostic, predictive and prescriptive paradigms in intelligent EV charging processes
     
Available Data

The available data is provided by the multiple physical facilities and digital services and includes historical, real-time and future data, such as:

  • Real-time and historical energy data from ECOs
  • Real-time and historical data from the testing labs
  • Historical data from the end-user’s testing facilities and pilots
  • Real-time and historical data from real TEF installations and simulations
  • Synthetic data from user and asset models in the DT
  • Historical data from past experiments and simulations
  • Future data from predictions, such as demand or weather conditions
     

Coming Soon!