Intelligent Earth Seminar: Maria Antonia Brovelli (Politecnico di Milano)

3 December 14:00
Seminar Room 1a, Doctoral Training Centre, 1-4 Keble Rd
mariaantoniabrovelli

Maria Antonia Brovelli is a renowned Professor at Politecnico di Milano, where she has built a distinguished career in GIS, Earth Observation, and sustainable development. With a Ph.D. in Geodesy and Cartography, she has held roles from researcher to Vice-Rector. Internationally, she serves as Vice President of ISPRS Commission IV and co-chair of the UN Open GIS Initiative. A leader in open-source GIS, her research has spanned geodesy, geospatial science, webGIS, Citizen Science, Big Geo Data, and geoAI. She has led and participated national, EU and ESA-funded projects and received major honors, including the OSGeo Sol Katz Award and ISPRS President’s Honorary Citation.

Geospatial Foundation Models Powering Sustainability and the SDGs

Geospatial Artificial Intelligence (GeoAI) represents the convergence of spatial data science and artificial intelligence, offering unprecedented opportunities to advance the United Nations Sustainable Development Goals (SDGs). By exploiting machine learning and deep learning, in particular transformer-based architectures, GeoAI enables sophisticated analysis of Earth Observation (EO) data, supporting applications from environmental monitoring to disaster response. In parallel, the field is also witnessing the rise of Structured State Space Models (SSMs), a new class of architectures that combine recurrence with efficient parallelization. Unlike Transformers, which rely on quadratic self-attention, SSMs can process very long sequences in linear time while retaining memory of past observations. Their potential is particularly relevant for geospatial applications where long EO time series and multi-sensor fusion require capturing dependencies across space and time without prohibitive computational costs. Other emerging approaches, such as recurrent–hybrid models and convolution-based alternatives, further indicate that the future of sequence modeling may extend beyond Transformers alone. By linking technological advancement with education and capacity building at global scale, GeoAI, Geospatial Foundation Models, and next-generation architectures such as SSMs are not only reshaping geospatial analysis but also contributing to a more equitable, informed, and sustainable future.