Project: Leonardo – Renewable energy production forecasting platform

Project duration: 1 January 2023 – 30 June 2025

Funding: The project is being implemented at a value of €1,788,586.03 (total project value: EUR 1,804,426.03 / eligible project value: EUR 1,788,586.03) and receives €992,882.45 in funding through the NextGeneration EU mechanism, using European funds.

Project description, purpose and planned results:

Leonardo is an AI-powered IoT platform for real-time, short-term and long-term forecasting of electricity production at photovoltaic power plants. By collecting data on key environmental parameters, using computer vision to monitor weather and applying advanced machine learning algorithms, it addresses the main obstacle to wider adoption of renewable energy sources (RES). It therefore has a significant indirect impact on reducing emissions and mitigating climate change.

Existing services forecast solar irradiance based on weather forecasts, local characteristics and the time of year. None of these solutions is reliable in the short term, as they do not capture short-term weather fluctuations. Computer vision systems can already detect current conditions, such as cloud movement, but are not integrated with other systems. Combining medium- and long-term forecasts with short-term data on current conditions enables machine learning models that more accurately predict the relevant environmental parameters determining electrical power output. This is the focus of the Leonardo project: developing a comprehensive, integrated platform that uses all relevant data to optimize production forecasting and maintenance of photovoltaic power plants.

The forecasting platform will provide more accurate forecasts of renewable energy production based on relevant environmental conditions, while correlating those forecasts with the actual condition and operation of the power plant. This makes it possible to detect deviations between actual and forecast production, quickly identifying potential problems and maintenance needs. We can thus ensure optimal operation and predictable energy production, a prerequisite for integrating larger amounts of renewable energy into the system. In these ways, such a forecasting platform indirectly contributes to reducing greenhouse gas emissions.