Hosted by the Faculty of Aerospace Engineering at TU Delft, the AI4REALNET consortium gathered on June 12-13 for its second meeting, a focused two-day event aimed at advancing artificial intelligence frameworks for critical network infrastructures. The meeting showcased the significant progress achieved since the consortium’s initial gathering and outlined strategic directions for the project’s next phase.

The first day opened with a review of major milestones, including discussions on developing a cohesive framework for AI applications. Participants explored advancements in AI building blocks, with an emphasis on integrating distributed reinforcement learning techniques.

A key highlight was a workshop on Human-Computer Interaction, where experts shared insights to enhance coordination and communication within the project. Discussions on validation and impact assessment stressed the importance of comprehensive evaluation protocols to ensure the AI systems’ reliability and effectiveness.

Day two focused on dissemination and engagement strategies, including plans for workshops and AI competitions. Members also presented experimental protocols for evaluating human-AI interactions, concluding with a steering committee session to set priorities for the upcoming months.

This meeting laid a solid foundation for AI4REALNET’s next steps. The consortium will reconvene soon to review progress and tackle new challenges.