Energyfish: AI-powered energy from river currents
RIVER (Resilient Infrastructure for Virtualization, Energy Generation, and Predictive Maintenance) develops a resilient, AI-driven infrastructure for hydrokinetic power plants, using the Energyfish as a case study. Its goal is to fundamentally improve the simulation, operation, and scaling of energy generation from river currents through digital twins, generative AI, and a modern IoT edge-cloud architecture. Real operational and environmental data are continuously collected, integrated with simulations, and used for precise representation of the installations. Building on this, AI-supported predictive maintenance enables a significant reduction in service interventions, operating costs, and downtime, while simultaneously extending the lifespan of the assets. Additional sensors support flood early warning, environmental monitoring, and water management. The technologies developed in the project are transferable to conventional small hydropower plants and provide the foundation for economically viable, sustainable, and socially accepted scaling in Bavaria and internationally.
The project makes a substantial contribution to the analysis, testing, and validation of digital twins for critical infrastructures, particularly energy systems and river and water systems. It advances the state of the art by combining automated test case generation, data-driven modeling, and simulation-based experiments to investigate system behavior, robustness, and failure modes of digital twins under realistic and extreme scenarios. A central contribution is the integrated consideration of simulation, test-adjacent environments, and real-world reference data to analyze and reduce the sim-to-real gap. Furthermore, the project generates empirically grounded insights, reusable software artifacts, and methodological guidelines aimed at improving the reliability, predictive accuracy, and decision-support capabilities of digital infrastructure digital twins.
15.01.2026 – 14.06.2026
