Tomorrow’s wind turbines will think for themselves

AAU Excellence participant Szymon Gres uses artificial intelligence and advanced mathematics to develop self-learning energy systems that can predict failures and increase lifetime.

How do you get a wind turbine to understand its own condition when the data it has available is limited and uncertain? Szymon Gres, Assistant Professor at the Department of Electronic Systems at AAU, is seeking for an answer to this question.

As a participant in the AAU Excellence programme, he does research on safe learning, where intelligent systems learn about themselves and their surroundings without compromising safety. 

"We want to learn something entirely new about a physical system without destroying it in the process. It may sound trivial, but it is actually one of the fundamental challenges in data-driven engineering," says Szymon Gres.

In search of faults no one can sense
In the wind energy industry, wind turbines are continuously monitored using sensors. But the measurements do not always give the full picture of the turbines’ health. Changing operating conditions, complex mechanical systems and limited access to data can make it difficult to detect faults before the critical failure of the system. 

So, Szymon Gres works with theories and algorithms that can reveal faults that are otherwise hidden in the available data.

"My vision is wind turbines that can understand their own dynamic state, anticipate failures before they occur, and collaborate across a wind farm to optimize both energy production and reliability."

An engineer in search for answers
His interest in research stems from a fundamental quest to understand how things work.

"I’m motivated by understanding and solving problems where theory meets engineering. Research offers a particular state of mind where a nerd like me has the freedom to fully investigate a problem, knowing that it can sometimes develop into a rabbit hole where no elegant, closed-form mathematical solution exists. But somehow, there is always a solution. It's fascinating!" says Szymon Gres. 

When different perspectives meet
In addition to his fascination with solving complex problems, an important driving force is the collaboration with colleagues and students.

"I’m passionate about working with colleagues who bring different and sometimes very opposing perspectives on science to the table, as well as with students whose curious and fresh minds often challenge established ideas."

From Rennes to Zürich and on to AAU
Research stays at Inria in Rennes in France and at ETH Zürich in Switzerland had a major impact on Szymon Gres' development as a researcher. In Rennes, he gained insight into the beauty of mathematics in engineering research, while his stay in Zurich introduced him to new theories and research areas and experiences in a top international research environment with high ambitions.

Today, he brings this experience to his work at AAU where he links his expertise with Learning and Decisions Lab’s research on secure artificial intelligence, regulation and data-driven decision-making. At the same time, he is helping to build AI:Wind-Lab, one of the nine AI:X Labs established in 2026.

Research with scientific impact in the EU
In five years' time, Szymon Gres' goal is to lead an international research group focusing on safe learning, fault diagnostics and forecasting for wind energy systems. 

Although his main priority will be developing theory for applications, he also plans to work toward establishing laboratory facilities for wind turbine research at AAU. He also hopes to obtain larger independent research grants that can anchor his group, and that the research group will have a central role in Europe across control engineering, machine learning and wind energy.

"My long-term ambition is to position AAU as a leading European centre for data-driven robustness and resilience in mechanical systems," he says.