FB 14: Participation in the Big Data from Space Conference, Riga, Latvia

Applicant: Yomna Eid
FB 14 – Geoinformatics, PhD Student

The Big Data from Space (BiDS 2025) conference is a biennial event organised by the European Space Agency (ESA), bringing together researchers, engineers, and decision-makers working at the intersection of Earth observation, artificial intelligence, and scalable data infrastructure. I attended the edition in Riga, Latvia, from 29 September to 3 October 2025, which felt particularly timely given how rapidly both the tooling and the science are shifting.

I contributed to the programme in two ways. First, a live demonstration of a machine learning workflow I developed with colleagues, combining R, Python, and Julia programming languages within a single Quarto notebook. The point wasn't to advocate for any one language, but to show how analysts can move fluidly between ecosystems without sacrificing reproducibility. People kept coming back to me throughout the day with follow-up questions, some of which pushed my thinking into territory I hadn't mapped out yet, and that kind of engagement is what makes presenting worthwhile. Second, I co-chaired a Birds of a Feather session on spatial data science across programming languages. The discussion went well beyond tooling: participants shared hands-on strategies for managing multilingual codebases, and we mapped out where current interoperability solutions still fall short. Hearing how practitioners at other institutions handle these problems day-to-day grounded my perspective in ways the academic literature rarely does.

Among the sessions I attended, two stood out. A presentation on a digital twin of European forests, integrating near-real-time Sentinel data with climate projections, directly mirrored challenges I face in my own work on multi-source data integration. The talks on foundation models for satellite imagery were equally thought-provoking, highlighting cases where classical machine learning still holds an edge while showcasing the expanding range of data modalities being brought into these models.

The informal side of the conference was equally valuable. Conversations over coffee with people working on adjacent problems led to a handful of connections I expect to follow up on concretely. BiDS 2025 sharpened my sense of where the field is heading, and gave me contacts and ideas that will directly shape my next steps. I'm grateful to the Santander Mobility Fund for making the trip possible.