This article provides a systematic understanding of the cutting-edge discipline of GeoAI through a comprehensive literature review. GeoAI is an emerging interdisciplinary field at the intersection of Geographic Information Systems (GIS), Artificial Intelligence (AI), and Geospatiotemporal Big Data (GBD). GIS has benefited significantly from the rapid advancement of AI, particularly through the integration of sophisticated AI algorithms into geospatial processing. As AI continues to evolve—especially with the emergence of Generative AI powered by high-performance computing and Large Language Models (LLMs)—there is an increasing need for the explicit handling of geospatial intelligence. This shift highlights the evolution of GeoAI from geospatial information to geospatiotemporal intelligence. It presents a valuable opportunity for GIS researchers to lead in AI innovation and contribute to addressing complex global challenges.

GeoAI, the integration of artificial intelligence and geographic information systems (GIS), is attracting attention as a new approach to solving social issues in urban, real estate, and architecture fields. In architecture, AI has been used for visualization and design automation, and its applications are expanding with advances such as large language models and image generation. As BIM (Building Information Modeling) becomes more widespread and standardized, integration with city models like CityGML is also progressing.

However, there are fundamental differences between GIS and BIM in terms of purpose, data precision, and privacy, which create challenges for data integration and practical use. Still, if unified, these data can enable broader applications of GeoAI for social problem-solving. This paper reviews the current situation, challenges, and future prospects of BIM and its integration with GIS, focusing on the new BIM-based building permit system to be introduced in Japan from spring 2026.

Artificial intelligence (AI) is increasingly transforming various domains, driven particularly by advancements in machine learning and deep learning, facilitated by large datasets and high-performance computing capabilities. Recently, Geospatial Artificial Intelligence (GeoAI), integrating AI techniques with georeferenced data such as satellite imagery and sensor observations, has emerged as a powerful tool for extracting geographic features, detecting environmental changes, and predicting spatial events. GeoAI is gaining prominence as an innovative solution across diverse applications, including smart cities, environmental management, and disaster response. However, significant challenges remain in addressing spatial heterogeneity, multimodal data integration, reproducibility, fairness, and transparency. Therefore, international standardization plays a crucial role in mitigating these challenges. This paper outlines GeoAI standardization initiatives within international organizations such as the Open Geospatial Consortium (OGC), International Organization for Standardization (ISO), International Telecommunication Union (ITU), and MLCommons, and discusses future directions toward improving interoperability, reliability, and scientific rigor in GeoAI technologies.

This study systematically reviews prompt engineering techniques in GeoAI, the intersection of geospatial information science and artificial intelligence, and verifies their effectiveness through application to walkability assessment. We developed “mapbotica,” a chat-based GeoAI tool built on large language models (LLMs), enabling spatial data queries and analysis through natural language. Through field experiments with 22 participants in Tokyo's Oimachi district, we collected 303 walkability evaluation records across 41 road segments, demonstrating that GeoAI-specific prompts, including spatial reference specification and Chain-of-Spatial flow, enable users without GIS expertise to perform spatial analysis with over 90％ accuracy. Our findings reveal non-linear relationships between physical infrastructure and subjective walkability assessments. The significance lies in establishing fundamental principles of GeoAI prompt engineering and demonstrating its practical applicability to participatory urban planning through quantitative evaluation.

Geospatial Artificial Intelligence (GeoAI) is an interdisciplinary field that integrates spatial information and artificial intelligence to enhance analysis and decision-making. This paper reviews the constituent technologies and recent advances in GeoAI, focusing on three phases of the spatial information processing cycle : data acquisition, processing, and dissemination. While practical applications have primarily advanced through deep learning-based image analytics, fully leveraging the natural language and multimodal capabilities of generative AI in geographic information science requires spatial ontologies and related infrastructure. Recognizing the gap between deep learning and generative AI, this study explores the potential applications of GeoAI from both challenge and outlook perspectives.

This paper provides an overview of recent trends in the social implementation of GeoAI, focusing on the application of Artificial Intelligence (AI) in Geographic Information Systems (GIS). It categorizes the use of AI technologies into two areas : Geospatial processing and the application of generative AI to GIS. The Paper examines case studies and research trends related to their real-world implementation. The use of AI in geospatial processing — such as machine learning, deep learning, and foundational AI models —has already led to the rapid development of numerous products and services. In contrast, research on the application of generative AI, such as Large Language Models (LLM), Autonomous GIS, to support GIS operations is still in progress. However, at present, LLM have not yetacquired the capability to perform complex geospatial processing tasks.

The JHS Geotechnical Evaluation and Prediction System leverages geospatial AI (GeoAI) to streamline ground evaluations for residential construction. By simply inputting a location, users can quickly assess ground strength, identify soft soil, determine improvement methods, and estimate costs. Its architecture, built on SuperMap's geospatial AI platform with SuperMap iServer 11i and PostgreSQL, delivers rapid geotechnical insights. A random forest model, trained with extensive JHS data and expert knowledge, achieves 70-80% accuracy in ground strength prediction. With nationwide coverage, the system significantly accelerates initial evaluations and decision-making. Future plans include enhancing the AI model, acquiring more detailed spatial data, and improving system processing to contribute to safer housing construction.

In recent years, artificial intelligence technology has been rapidly developing, and the use of artificial intelligence technology is widely spreading in the field of geospatial information science, called GeoAI. This paper organizes satellite remote sensing data as GeoAI from the perspective of Big Data and Data Science, introduces an example of GeoAI using remote sensing data, and discusses recent trends and prospects. Satellite remote sensing data as big data mainly divided into optical satellite data and microwave satellite data, for example, Synthetic Aperture Radar (SAR) satellite data, and are recently emerged frequent observations by small satellites, and also free. Data science has progressed with machine learning and deep learning methods. As an example, this paper take the study of Xu et al. (2024), which is used huge amounts of satellite remote sensing data and deep learning to create crop map. As a recent trend, multimodal analysis and foundation model are mentioned.

This paper introduces the activities and underlying objectives of the Young members' special interest group. Since its establishment in 2016, our subcommittee has aimed to promote interaction among young members, including early-career researchers, practitioners, and students, and further, to broader GIS community. Its main activities include organizing sessions at annual conferences, hosting student research presentations, and conducting surveys on GIS education or academic engagement. Moving forward, the subcommittee will continue to promote our initiatives in line with its founding objectives.

International Digital Earth Applied Science Research Center (IDEAS), Chubu University, has been accredited by the Minister of MEXT as “Joint Usage/Research Center for Digital Earth to Address Emerging Complex Systemic Problems” and is actively promoting research under this framework. This article outlines the research framework of the center with particular emphasis on the themes of the competitive research projects at its core. It provides detailed descriptions of nine research themes, including the Special Category “Exploring the Sustainability of Social-Ecological Systems,” as well as two thematic categories : “Category 1 : Integration of Technical Components for Digital Earth” and “Category 2 : Digital Earth Applications to address Complex Systemic Problems”. This article provide an overview of the academic foundations and structure of “Digital Earth”, along with a comprehensive perspective on the complex systemic problems that serve as the research focus of the center.

As the nation's foremost map-making organization, the mission of Geospatial Information Authority of Japan (GSI) is to provide accurate positions and fresh maps to the public, through four roles of “measuring,” “portraying,” “safeguarding,” and “conveying.” GSI is working on the development, expanded use, and social implementation of geospatial information, which contributes to national economy development, disaster risk reduction, and national security. This paper introduces some of the major recent initiatives by GSI concerning the development and utilization of geospatial information.

Project PLATEAU, launched by Japan's Ministry of Land, Infrastructure, Transport and Tourism in 2020, aims to advance urban development through the digital transformation and open data sharing of 3D city models. Involving over 100 organizations, including traditional geospatial firms and startups, the initiative spans various sectors such as urban planning, disaster prevention, and energy management. This paper highlights the contributions of Asia Air Survey Co., Ltd. to PLATEAU, detailing the development of 3D city model specifications, data creation, visualization, distribution and use cases. While challenges remain—such as establishing regular updates for 3D models and refining the regulatory framework for AI-generated data—ongoing collaboration and technological development are essential for the successful implementation of 3D geospatial information in Japan.

This study reviews recent R packages for spatial statistical modeling, focusing on scalable alternatives to Gaussian processes (GPs), which are limited by high computational cost. We categorize key approximation strategies into low-rank basis methods, covariance tapering, and sparse precision matrix approaches such as the SPDE method. Among them, the sdmTMB package, which supports generalized linear mixed models, stands out for its computational efficiency, flexibility, and ease of use. We demonstrate its practical utility through a case study on fish distribution data, highlighting its ability to model spatiotemporal variation and terrain-constrained processes.

Accurate human mobility data are crucial for urban planning, transportation management, and disaster preparedness, yet traditional data sources often face limitations due to privacy issues, cost, and data sparsity. Recently, large language models (LLMs) have emerged as promising tools for generating synthetic human mobility data. This paper reviews state-of-the-art LLM-based methodologies, categorized into training-from-scratch, hybrid fine-tuning, and prompt-constrained autoregression, highlighting their respective strengths and limitations. We further compare various spatial tokenization approaches—including grid-based, road network-based, POI-based, coordinate quantization, and hierarchical encoding—and examine practical applications such as event-driven forecasting, missing-data imputation, and interactive scenario simulations. Finally, we discuss remaining challenges and future directions concerning spatial consistency, computational efficiency, evaluation standardization, multi-resolution modeling, and human-AI collaborative frameworks. We provide practical insights to advance realistic yet privacy-preserving human mobility simulations.