Politecnico di Milano, in Italy, welcomed the 17 partners of the AI4REALNET project, who gathered for the third General Assembly on the 3^rd and 4^th of June.

This was a two-day event that counted on over 40 participants, in person and online. The purpose of the event was to present a point of order in all the work that has been developed, showcase results and progress, as well as provide opening discussions to decide the next steps in the human-AI cooperation in the operation of critical infrastructures.

The first day started with a quick overview of all the achievements accomplished by the different work packages in the last year. Afterwards, a cross-discussion between partners was directed towards the possible adaptation of some of the results to different use cases other than the ones targeted in the project: power grid, railway, and air traffic management.

During the afternoon, and on the second day, important topics were discussed, such as the interoperability of the AI-friendly digital environments, the different interactive parts between humans and artificial intelligence, AI validation campaigns, and protocols. As a result of these discussions, strategic measures were outlined for the project’s next phase, including plans of action for the upcoming months.

This was the third time the consortium had gathered since the beginning of the project, with the next General Assembly scheduled for the beginning of 2026.