AgenticCity-Bench tracks urban changes with multiple data in 16 cities
AgenticCity-Bench: Longitudinal Multimodal Observations of Urban Dynamics with Agentic Reasoning
Artificial Intelligence
Summary
Cities are complex systems where buildings, people, and the environment constantly interact. Existing tools don’t easily combine different types of data over time to understand these interactions. The authors created AgenticCity-Bench, a big dataset with many kinds of information from 16 cities worldwide collected from 2000 to 2025. This dataset helps computer programs learn how cities change and make decisions about planning and resources. It combines data on city functions, activities, and environment all indexed by location and date for easier study.
What this means in practice
- •For urban planners: Use long-term multimodal city data to improve forecasting and resource allocation in urban development projects.
- •For environmental monitoring teams: Integrate diverse urban and environmental data to better track and respond to changes in city ecosystems over time.
- •For smart city technology developers: Build AI systems that reason about city states and plan actions using unified multimodal urban datasets.$Commercial implications: Enables commercial AI products for city management by providing comprehensive data for decision-making automation.
Authors
Xiaofei Yu, Kailai Sun, Yanyu Mao, Baoshen Guo, Heye Huang, Yongcai Wang, Baotong Mi, Alok Prakash, Shenhao Wang, Jinhua Zhao
Abstract
City systems are shaped by interactions among infrastructure, human activity, and environmental conditions. Existing earth observation and urban benchmarks are limited in unified modeling and in capturing their joint dynamics. We introduce AgenticCity-Bench, a longitudinal multimodal benchmark for multi-layer unified state modeling and agentic city task reasoning. The benchmark covers 16 cities in 12 countries across six continents from 2000 to 2025, and includes 18 modalities spanning urban function, activity, and environmental state. It organizes heterogeneous data sources around a common spatial reference and calendar index while preserving spatial and temporal coverage, provenance and tile-level availability. On this foundation, AgenticCity-Bench supports urban cross-modal translation together with diverse agentic city tasks involving forecasting, planning and resource analysis. This study provides a unified benchmark for studying the transition from multimodal urban and earth observation to joint understanding and agentic city decision-making. The dataset is available at https://anonymous.4open.science/r/AgenticCity-Bench.