On 11-12 March 2026, the BRAID consortium met in Kaiserslautern for its first in-person workshop. Researchers from KIT, RPTU, TU Dortmund, GSI and inomed came together to launch the scientific exchange at the heart of the project and to discuss how common AI methods can be developed across very different application domains.
The workshop was designed as a kick-off meeting: a chance to align ideas, define shared priorities, and identify where the different domains face similar challenges when working with irregular high-dimensional data.
Across two days, the discussions focused on central BRAID themes such as physics-aware machine learning, learned locality, dynamic dimensionality reduction, algorithmic constraints, and common software infrastructure. The meeting also made clear that, despite the differences between the domains, there is strong common ground in the underlying reconstruction problems. At the same time, the workshop already helped crystallise several first proof-of-concept directions that will now be explored further in the project.
Just as importantly, the workshop provided time for informal exchange. The excellent local organisation by the RPTU team, including a joint dinner in Kaiserslautern, created a very good setting for discussions beyond the formal programme and helped turn the meeting into a genuine starting point for collaboration across the consortium.
With this first workshop, BRAID has taken an important step from project planning into active scientific exchange. The ideas discussed in Kaiserslautern will now feed into the next phase of method development and cross-domain collaboration.