Urban environments, characterized by their complex, multi-layered networks encompassing physical, social, economic, and environmental dimensions, present significant challenges for sustainable urbanization. These challenges, ranging from traffic congestion and pollution to social inequality, call for advanced technological interventions. The technology innovation in big data, artificial intelligence, urban computing, and digital twins have laid the groundwork for sophisticated city simulation. However, a gap persists between these technological capabilities and their practical implementation in addressing urban issues because they often fall short of capturing the complex and subtle human behaviour in urban space. The recent advance in large language model (LLM) agents shows emergent abilities of human-like behaviour simulation, presenting important opportunities for characterizing human behaviour in urban studies. This paper provides a comprehensive review on the recent literature about the technology development of urban computing, digital twins, LLM agents and beyond, as well as the interdisciplinary studies on complex urban system and agent-based modeling. Moreover, we conceptualize a novel Urban Generative Intelligence platform that grounded LLM agents in simulated urban environment. The UGI platform allows LLM agents to operate within a textual urban environment emulated by city simulator, interact through a natural language interface, offering an open platform for diverse intelligent and embodied urban tasks. Such platform unleashes the power of LLM agents for complex urban system simulation, providing a novel approach to understand and manage urban complexity.
Xu et al. (Fri,) studied this question.