Developing GeoAI methods that enable machines to understand, represent, learn from, and reason with spatial information for complex real-world applications. This direction advances spatial representation learning, data and knowledge engineering, multimodal spatial information extraction, and spatial reasoning across different spatial data modalities, including maps, imagery, trajectories, networks, text, sensors, and urban knowledge graphs.
Advancing data-driven intelligence for transportation systems, mobility services, and large-scale network analytics. This research focuses on sensing, modeling, predicting, and optimizing movement in space, including traffic information extraction, multimodal data fusion, real-time forecasting, routing, navigation, network analytics, and intelligent location-based services.
Applying spatial intelligence to understand and address pressing urban and societal challenges by transforming heterogeneous urban data into actionable insights for cities, communities, and decision makers. This direction integrates big spatial data, domain knowledge, computational models, and urban AI to support scientific discovery and decision-making in areas such as resilience, health, population dynamics, logistics, city services, sustainability, and public policy.
Designing optimization methods, computational frameworks, and production-level GeoAI systems for real-world services and decision-making. This direction focuses on graph and network optimization, routing, resource allocation, service-area design, spatial data sensing and production, and human-in-the-loop decision support. It bridges algorithmic development and operational implementation, with strong emphasis on scalability, interpretability, reliability, and deployment through industry and government partnerships.
