AutoML, Benchmarking and Open Science Infrastructure for Earth Observation 2026

The workshop location in Ljubljana is well-chosen for European researchers. The topic of open science infrastructure for Earth Observation is highly relevant and timely for the field.

The workshop’s focus on the intersection of AutoML and Earth Observation infrastructure appears well-conceived, particularly given the increasing complexity of multimodal satellite archives and self-supervised learning models. The involvement of key institutions like the University of Ljubljana, Leiden University, and the Jožef Stefan Institute suggests a strong academic foundation. The mention of FAIR data principles and standardized metadata indicates a serious commitment to reproducibility, which is crucial for the development of reliable AI assistants and agentic systems in this domain. However, the fact that the full list of keynote speakers is not yet announced leaves some uncertainty about the depth of industry perspectives that will be included. The reliance on abstract concepts like workflow optimization and empirical performance modelling is promising, but it remains to be seen how these will be applied in practical benchmarks. The contact information is available for those with specific questions, which is a good sign for transparency. Overall, the initiative seems valuable for researchers looking to understand the next generation of automated analysis pipelines, provided the final program delivers on the breadth of topics suggested by the organizers.

While the scientific scope regarding automated Earth Observation models is clearly defined, the practical details for attendees remain somewhat vague. It is not explicitly stated whether registration fees apply, how the keynote speaker list will be finalized, or what the specific submission deadlines are beyond the general call. For someone considering attending or submitting a paper, these operational terms need to be clarified. The contact email is provided, but it is unclear if a dedicated registration portal exists or if all inquiries go through the organizers. Until the full speaker list and venue logistics are confirmed, potential participants should verify all administrative requirements independently to avoid any last-minute surprises regarding their participation.

The organizational structure appears solid, with clear involvement from universities and research institutes like the University of Ljubljana. The integration of FAIR data principles and controlled vocabularies suggests a rigorous approach to open science infrastructure for machine learning applications in Earth Observation.

The workshop highlights the essential role of AutoML in Earth Observation research. The emphasis on automated pipelines seems very promising for future scientific workflows.