Delivered on October 26, 2026
Understanding Cities with GeoAI: How Human Flow Data x AI Will Transform the Future of Urban Analysis
Speakers:
Shun Nakayama, Specially Appointed Professor, Faculty of Network and Information Science, Senshu University
Affiliation and position information is as of the time of distribution
Our guest this time is Professor Shun Nakayama, a specially appointed professor in the Faculty of Network and Information Science at Senshu University. Specializing in geographic and spatial information, Professor Nakayama researches the relationship between human flow data and urban structure. We will ask him about the current state of the rapidly evolving "GeoAI." How can "GeoTransformer," which applies the Transformer technology supporting generative AI to spatial analysis, capture the complex structure of cities and human movement, and connect this to simulations for redevelopment? Furthermore, starting from the question, "Can AI read maps?", we will discuss the future of GeoAI from both research and social implementation perspectives, covering the rapid evolution of GeoAI in just a few months, the importance of traditional GIS and geographical knowledge, and the reskilling required of business professionals.
＝＝＝＝
The Background of Professor Shun Nakayama, Specially Appointed at Senshu University, and the Origins of His Research on Urban Structure and Human Flow Data
Professor Shun Nakayama, Specially Appointed at the Faculty of Network and Information Sciences, Senshu University, specializes in geographic and spatial information, and is engaged in research to elucidate the relationship between urban structure and human flow data.
Professor Nakayama's doctoral dissertation focused on the relationship between urban density, floor area ratio, and decarbonization. After obtaining his degree, he changed his research theme and is now developing research on how urban density and floor area ratio relate to people's movement patterns while handling human flow data. He heads a relatively new research laboratory established at Senshu University, and is working with undergraduate students to create research results and cultivate talent through trial and error.
The Innovative Nature of "GeoTransformer," Applying Generative AI Technology to Spatial Analysis
In recent years, the development of advanced spatial processing technologies combining artificial intelligence has been rapidly progressing in the field of geospatial information.
The application of "GeoTransformer," which applies Transformer—a foundational technology that underpinned the rapid advancements in generative AI—to spatial data processing is advancing. This allows for the capture of complex interactions between geographically dispersed areas within a city, which was difficult with conventional methods, enabling a precise understanding of urban structure. This allows for a clear explanation of how complex urban structures trigger interactions and, as a result, how much pedestrian flow occurs in specific areas.
Urban Redevelopment and Pedestrian Flow Simulation: Application to the Placement of High-Rise Condominiums and Commercial Facilities
Mobility analysis combining GeoTransformer and pedestrian flow data provides a powerful simulation foundation for decision-making in urban planning and redevelopment projects.
It can model how pedestrian numbers and movement patterns change throughout a city when new commercial facilities or high-rise condominiums are placed in a redevelopment area. As a practical tool for predicting pedestrian flow changes after policy implementation, it supports evidence-based urban development planning.
Definition and Evolution Speed of "GeoAI": Inter-Layer Interaction and the Current State of Map Interpretation
GeoAI (Geospatial Artificial Intelligence) has evolved beyond mere image analysis into a technology that comprehensively understands the multi-layered structure of geographic information.
Unlike conventional satellite image analysis, the essence of GeoAI lies in its advanced learning and analysis of the diverse information layers and their complex relationships within GIS. Technology is being updated every few months, starting with natural language-based GIS processing, and with proper training, even the latest AI can adequately interpret map information and spatial relationships. The application of Transformer technology is becoming mainstream in global academic conferences, and the pace of technological evolution is extremely rapid.
The Importance of Universal Classical Geography and GIS Knowledge and its Integration with the Latest Technology
When dealing with GeoAI, a cutting-edge technology, it is essential not only to keep up with the latest algorithms but also to have a solid foundation in classical geography and GIS theory.
The question of how to divide urban space into a mesh-like structure is a fundamental theme that has been debated for many years in classical geography. The latest GeoAI technology can only be properly utilized thanks to accumulated knowledge from existing GIS layer management and historical spatial theories. Rather than discarding past knowledge, integrating historical geographical questions with the latest models enables more fundamental urban analysis.
GeoAI Reskilling and Industry-Academia Collaboration Challenges for Business Professionals
Reskilling among management and practitioners is crucial for the social implementation of GeoAI in business settings and for making appropriate decisions.
A gap exists in the perception and understanding of GeoAI among practitioners, hindering appropriate decision-making. Educational programs to correctly grasp the mechanisms and potential of the technology, as well as the promotion of retraining for working professionals through academia such as universities, are expected.
Summary
Professor Shun Nakayama of Senshu University's research on "GeoAI" and "GeoTransformer" visualizes the complex relationship between human flow data and urban structure, representing a cutting-edge approach that will revolutionize urban planning and redevelopment simulations.
The importance of integrating cutting-edge AI technology with the fundamental concepts of GIS and the knowledge of classical geography, cultivated over many years, rather than blindly trusting or relying on it, has been demonstrated. Furthermore, reskilling for business professionals to keep up with rapidly changing technology and promoting industry-academia collaboration are key to the successful social implementation of GeoAI. The fusion of human flow data and GeoAI will powerfully drive the creation of sustainable and prosperous cities for the future.