On the first day, we kicked off with an introduction to IEA Wind TCP from Amy Robertson (NREL), followed by a key-note presentation “Harnessing Digital Twins: Pioneering the Future of Industry Digitalization” from José Valenzuela del Rio, Siemens Foundational Technology. The recording is available here and below:

After an overall task update from our Co-Operating Agent Sarah Barber, we started with the four working group workshops, which each lasted about one hour. In the WG1 (Data) workshop, we reported our progress on the IEA Technical Report “Evolving the wind energy sector towards frictionless and sustainable data usage”, the Wind Energy Domain Ontology, the TechnoPortal, our Recommended Practice: Ontology creation, publication, and maintenance for wind energy domain experts, and our planned Joint Best Practices for the Classification and Valuation of Digital Twins in Wind Energy. We then discussed high-potential Digital Twin use cases, their classification and valuation.

In the WG2 (Culture) workshop, we presented our recent progress on analysing the results of the IEA Wind Task 43 Culture Questionnaire, a draft of which was also presented at WESC2025 on June 26th, 2025, and is available here. After a Mentimeter survey, in which participants expressed some concerns about the effects of generative AI on organisational culture, we had an update on the skills mapping activity, in which we are assessing the skills needed with and without using the IEA Wind Task 43 WRA Data Model for wind resource assessment activities. The result will be a recommendation for organisations wanting to optimise hiring and training in the digital era, as well as a general framework for repeating the skills mapping to other areas. We ended with an update on the EU-funded DigiWind project from Nikolay Dimitrov from DTU Wind.

In the WG3 (Coopetition) workshop, we presented our progress on the WeDoWind open innovation ecosystem, which we also presented at WESC2025 on June 26th, 2025 (available here), and on the IEA Wind Task 43 Data User Group, which was launched in December 2024 and aims to develop tools for dealing with SCADA alarms. We then ran a workshop about the challenges and opportunities of open-source software in wind energy, coming to the conclusion that community building and long-term maintenance are key topics that IEA Wind Task 43 could contribute to. We’ve set up a WeDoWind group here, and will be in touch with interested parties soon!

In the WG4 (Use Case Demonstration) workshop, we first gave a working group status update, within which a draft deep dive use case or digitalisation activity process and value assessment framework was introduced. Then we presented progress on four currently active deep dive use cases. New results include: 1) WRA data model team released latest version 1.4 to include floating lidar file format and accessible here; 2) Risk-based maintenance on blade leading edge erosion team optimized the repair and inspection strategy out of over 30 candidates in terms of cost, which will be reported in an upcoming journal article; 3) Risk-based maintenance on gearbox reliability team developed a gearbox damage state taxonomy, a vulnerability framework for gearbox, and a Cook’s method elicitation exercise; and 4) Digital-twins as decision support tools team presented case study results through the collaboration with Fraunhofer by using a 8 MW turbine model. Three of these four deep dive use cases were also presented at WESC 2025.

This was followed by an overlaps and synergies workshop, in which several conclusions were drawn about future activities:

• We need to keep working on the Wind Energy Domain Ontology to improve the FAIRness and AI-readiness of wind energy data.
• We should look more directly at funding possibilities to enhance our work.
• We should develop partnerships with organisations such as the ESIG O&M User Group, the planned IEA Task on AI in Wind Energy, the Digital Twin Consortium, and the Research Data Alliance.
• We should work more directly with other IEA Tasks, in particular on topics such as lidar, lifecycle and environmental impacts.
• We should remain active and keep organising events in which we get valuable inputs from the industry.

On the second day, we started with an introduction to WeDoWind and a workshop in which some participants tested the use of Renku for future WeDoWind challenges and some participants created a new WeDoWind space and gave us feedback on the functionality. Renku is an open-source platform—created by the Swiss Data Science Center at EPFL/ETH Zürich—for doing reproducible, collaborative data science in a single, cloud-ready workspace. It links version-controlled code (Git/GitLab), data sets and containerised compute environments, records the full lineage of every result in a built-in knowledge graph, and launches on-demand Jupyter, RStudio or VS Code sessions so teams can experiment without local installs. If you are interested in testing Renku and WeDoWind, join our Open Data Exploration space here. This session included a report from Yu Ding about his WeDoWind Data Science for Wind Energy Space, which you can access here.

This was followed by two presentations about generative AI in wind energy:

• “Intelligent digital twins with wind energy domain knowledge grounding” by Yuriy Marykovskiy, Eastern Switzerland University of Applied Sciences (recording here)
• “Automated question-answering for interactive decision support in operations & maintenance of wind turbines” by Joyjit Chatterjee, EPAM Systems (recording here)

The presentations are embedded below:

We closed with a panel discussion about “The opportunities and challenges of generative AI in wind energy”, moderated by Sarah Barber. the participants, Joyjit Chatterjee from EPAM Systems, Yuriy Marykovskiy from the Eastern Switzerland University of Applied Sciences, Gareth O’Brien from Microsoft, Saumya Sinha from NREL and José Valenzuela del Río from Siemens Foundational Technologies, talked about the challenges and opportunity of generative AI in wind energy. the general consensus was that generative AI is here, if we like it or not, and the wind energy sector needs to adapt. We discussed concerns about the energy needs of generative AI as well as ethical issues. We came to the inclusion that common data models such as the Wind Energy Domain Ontology are indispensible for the future of wind energy.

Find out more about IEA Wind Task 43 here, or get in touch with Sarah Barber at sarah.barber@ost.ch or Shawn Sheng at shawn.sheng@nrel.gov.