Skip to main content


The Digital Twin solution for AI-driven hydropower energy forecasting


Developed by
Business Category
Technical Category
AI services Machine learning

River discharge rules energy production for Hydropower plants.
Prediction of water resources for the next day, month, season, challenges every energy producer and trader.

Such knowledge supports optimal energy production, avoiding wastes (underestimation) or empty reservoirs (overestimation). 

Plant managers convert this forecast into energy by applying plant specific (mostly confidential) efficiency factors.

Sound forecasting leads to better market financial strategies (energy trading in advance to match most favourable market conditions).

Solving this challenge requires routinely complex forecasts models that describe physical behaviour of the catchment and the rainfall runoff transformation (eventually including snow component) to be set up individually for every plant and fed with a huge amount of local data different from plant to plant.

SmartRIVER is a Digital Twin of the plant catchment, relying on worldwide available forecasts, big open climate, geospatial and satellite data, and AI at the core, for efficient energy forecasting, with major advantages over complex models:

-    No specialist hydrologist skills, no physical model to tune.
-    Easy integration of upstream services and public datasets (e.g., Copernicus Climate Data Store) 
-    Lightweight cloud deployment, just a web browser is required to users.

SmartRIVER is provided as  Software as a Service (SaaS) and can be activated for every hydropower plant of interest to support water resources management and energy production.

Please read the Prototype manual to login and  access the service

On the github page also the example datasets beyond the prototype are provided ("Data" folder)


SmartRIVER prototype has been developed in the 1st Open Call of I-NERGY (This project has received funding from the European Union's Horizon 2020 research and innovation programme within the framework of the I-NERGY Project, funded under grant agreement No 101016508)