Closure of the IA4TES project

• Spanning three years (2022–2024), the project investigates solutions provided by various artificial intelligence technologies in the electric sector. Supported by the Next Generation EU initiative and the R&D Missions in AI Program, the consortium was formed, led by Iberdrola España and comprised of companies, research centers (including BCAM), and universities.
• Researchers Santiago Mazuelas (BCAM – Ikerbasque) and Vincenzo Nava (BCAM-Tecnalia),participated in the project, along with 17 other entities.

The IA4TES project, developed by a consortium led by Iberdrola España and composed of companies, research centers (including BCAM), and universities, has concluded after three years. Researchers Santiago Mazuelas (BCAM – Ikerbasque) and Vincenzo Nava (BCAM-Tecnalia) were part of the project.

IA4TES is a research project under the framework of the R&D Missions Program in Artificial Intelligence and supported by the Next Generation EU initiative and the R&D Missions in AI Program. The project was ranked third among five selected in Spain, earning 63 points and receiving total funding of over €12.5 million.

This initiative tackled key challenges in energy management through artificial intelligence and advanced mathematical models. Collaborative efforts among all consortium partners significantly contributed to developing methods for predicting energy load and pricing, essential for optimizing generation capacity planning, balancing supply and demand, and reducing energy trade costs.

Problem Significance
In the context of rapidly evolving energy systems and smart grids, precise forecasting of load and prices has become an essential tool to address uncertainties inherent in energy demand and the financial variables associated with it. Moreover, advancements in this field have a direct impact on sustainability by reducing energy waste and the costs associated with its procurement.

Key Achievements
Within BCAM, the IA4TES project has achieved notable results in both academic and industrial fields:

• Impactful Publications: Research such as “Probabilistic Load Forecasting Based on Adaptive Online Learning” was published in high-ranking journals like IEEE Transactions on Power Systems (2021).
• Awards and Recognitions: The work won the Best Applied Contribution in Statistics Award by the Spanish Society of Statistics and Operations Research (SEIO) and BBVA Foundation in 2022.
• Industrial Collaboration: Collaboration with Iberdrola's financial department was crucial in defining research objectives and the most effective methodological strategies.
• Media Coverage: The project’s significance was highlighted in national media outlets like Antena 3, Diario ABC, and El Mundo, as well as in an article in the bulletin of the Royal Spanish Mathematical Society (RSME).

Mathematical and Computational Methods
The team developed predictive methods that not only assess uncertainties associated with energy variables but also adapt to changes in consumption and cost patterns (concept drift). These efficient and flexible algorithms generate predictions based on hundreds of models updated with real-time data, achieving significantly improved performance across various scenarios.

BCAM’s involvement and contributions focused on the following activities:

• Activity 3.16: Structural Health Monitoring of Offshore Wind Turbines
• Activity 3.17: Estimation of Lifespan of Offshore Turbine Components
• Activity 4.6: Multidimensional Demand Forecasting
• Activity 6.11: Forecasting/Price or Demand Spikes in Energy

Impact and Conclusions
The project’s closing meeting was held during a two-day event on December 3 and 4 at the Iberdrola Campus in San Agustín del Guadalix. This event provided an opportunity to share achievements and thank everyone who contributed to the work over these years.

The results obtained within the IA4TES project have not only improved available prediction tools but also established a benchmark in applying artificial intelligence to the energy sector. We conclude this stage with the satisfaction of having delivered innovative solutions that contribute to technological advancement and environmental sustainability.

Consortium Partners

The consortium included two major national leaders: Iberdrola in energy and Indra-Minsait in digitalization with AI specialization. Additionally, nine SMEs with expertise in AI and energy or other sectors contributed, including Ariadna Grid, Balantia, Baobab Soluciones, Barbara, Bedrock, Efficiency Program, Fexidao, RatedPower, Singlar Innovation, Stemy Energy, and Lokinn.

Four research institutions also participated: Tecnalia, the Institute of Knowledge Engineering, Vicomtech, and BCAM. Lastly, three universities—the University of Granada, University of Salamanca, and Polytechnic University of Madrid—played crucial roles in the project.