TUM BGDM at the AI4AGRI Summerschool of the Romanian Excellence Center for Artificial Intelligence

This week, the 2025 AI4AGRI Summer School of the Romanian Excellence Center for AI in Agriculture took place in Brasov, Rumania. It is a very traditional and successful summerschool. This year, more than 45 participants listened to talks from European experts and shared the current state of the art in Artificial Intelligence for Agriculture with participants.

Prof. Martin Werner gave a half-day tutorial on parallel processing with Python and it was very well received. In times where the datasets, especially in agriculture, face severe Big Data integration challenges from two perspectives: the size and amount of Earth Observation data comes togehter with the very high number and fragmentation of the agricultural parcels in Europe.

Beyond learning a bit about the Dask parallel programming framework for Python, the students ran through a practical example of organizing OpenStreetMap data for quick access using Spatial Hashing and Bloom Filters.

Beyond science, this summerschool has been an awesome experience bringing together not only students, but as well scientists from across Europe in the area of AI4AGRI. We are deeply impressed by the results that people obtain from free Copernicus data.

All educational material is freely available on Zenodo as open education, see https://zenodo.org/records/16456370.