ESSI1.4 | Advances in Geospatial Artificial Intelligence for large-scale, regional and continental mapping

ESSI1.4

EDI
Advances in Geospatial Artificial Intelligence for large-scale, regional and continental mapping
Convener: Nouri Sabo | Co-conveners: Michael Tischler, Alexandre Hippert-FerrerECSECS, E. M. FarellaECSECS, Ewelina Rupnik

Geospatial artificial intelligence (GeoAI) has gained popularity for creating maps, performing analyses, and developing geospatial applications that are national, international or global in scope, thanks to its capacity to process and understand large geospatial data and infer valuable patterns and information. Rapid geo-information updates, public safety improvement, smart city developments, green deal transition as well as climate change mitigation and adaptation are among the problems that can now be studied and addressed using GeoAI.
Along with the acceleration of GeoAI adoption, a new set of implementation challenges is ascending to the top of the agenda for leaders in mapping technologies. These challenges relate to “operationalizing large GeoAI systems”, including automating the AI lifecycle, tracking and adapting models to new contexts and landscapes, temporal and spatial upscaling of models, improving explainability, balancing cost and performance, creating resilient and future-proof AI and IT operations, and managing activities across Cloud and on-premise environments.
This session aims to provide a venue to present the latest applications of GeoAI for mapping at national, international and global scales as well as their operationalization challenges. The themes of the session include, but are not limited to:
· Requirements of GeoAI methods for national mapping agencies, their relationship with industrial/commercial stakeholders, and the role of national agencies in establishing GeoAI standards.
· GeoAI interoperability and research translation.
· Extracting core geospatial layers and enhancing national basemaps from multi-scale, multi-modal remote-sensing data sources.
· Large-scale point cloud analysis for use cases in infrastructure development, urban planning, forest inventory, energy consumption/generation modeling, and natural resources management.
· Measuring rates and trends of changes in landscape patterns and processes such as land-cover/land-use change detection and disaster damage proxy mapping.
· Modernizing national archives, including geo-referencing, multi-temporal co-registration, super-resolution, colorization, and analysis of historical air photos.

Geospatial artificial intelligence (GeoAI) has gained popularity for creating maps, performing analyses, and developing geospatial applications that are national, international or global in scope, thanks to its capacity to process and understand large geospatial data and infer valuable patterns and information. Rapid geo-information updates, public safety improvement, smart city developments, green deal transition as well as climate change mitigation and adaptation are among the problems that can now be studied and addressed using GeoAI.
Along with the acceleration of GeoAI adoption, a new set of implementation challenges is ascending to the top of the agenda for leaders in mapping technologies. These challenges relate to “operationalizing large GeoAI systems”, including automating the AI lifecycle, tracking and adapting models to new contexts and landscapes, temporal and spatial upscaling of models, improving explainability, balancing cost and performance, creating resilient and future-proof AI and IT operations, and managing activities across Cloud and on-premise environments.
This session aims to provide a venue to present the latest applications of GeoAI for mapping at national, international and global scales as well as their operationalization challenges. The themes of the session include, but are not limited to:
· Requirements of GeoAI methods for national mapping agencies, their relationship with industrial/commercial stakeholders, and the role of national agencies in establishing GeoAI standards.
· GeoAI interoperability and research translation.
· Extracting core geospatial layers and enhancing national basemaps from multi-scale, multi-modal remote-sensing data sources.
· Large-scale point cloud analysis for use cases in infrastructure development, urban planning, forest inventory, energy consumption/generation modeling, and natural resources management.
· Measuring rates and trends of changes in landscape patterns and processes such as land-cover/land-use change detection and disaster damage proxy mapping.
· Modernizing national archives, including geo-referencing, multi-temporal co-registration, super-resolution, colorization, and analysis of historical air photos.