AI integrated framework for intelligent geospatial handling and robust operation in MultiGIS applications

MultiGIS provides precise and real-time spatiotemporal data analysis, which greatly benefits various applications including environmental monitoring, sustainable development, resilience-building, and urban planning. Despite wider adoption of MultiGIS applications, there are significant challenges remaining, notably relating to data collection, generation, storage, and real-time data processing from a set of diverse heterogeneous sources within MultiGIS environments. Due to limited data availability in most of the MultiGIS applications, it is necessary to generate synthetic data to enrich MultiGIS representations. Existing GIS-backed geostatistical AI tools can address some of these challenges but are limited with spatial autocorrelation and lack plug and play integration with existing tools as well as assurance of transparency, privacy, fairness, regulatory compliance of AI-enabled MultiGIS applications.

AI4MultiGIS offers an integrated framework for MultiGIS data generation and management, which aims to enhance the overall GIS capabilities for optimised processing chain of MultiGIS applications. The project will deliver innovative solutions to collect and consolidate the data from diverse sources. It will develop techniques for automated outlier detection for reliable data processing, real-time spatiotemporal data processing for informed decision making and plug and play capabilities using a cloud-integrated AI platform for multi-modal data integration. AI4MultiGIS considers assurance of responsible AI within the AI-enabled MultiGIS applications through development and implementation of policy and best-practice guidelines. Finally, AI4MultiGIS is evaluated through two pilot case studies to demonstrate their effectiveness and wider adoption of AI4MultiGIS in different MultiGIS applications beyond the project.