Leveraging Generative AI for Urban Digital Twins: A Scoping Review on the Autonomous Generation of Urban Data, Scenarios, Designs, and 3D City Models for Smart City Advancement
Haowen Xu, Femi Omitaomu, Soheil Sabri, Sisi Zlatanova, Xiao Li, Yongze Song
TL;DR
This review assesses how Generative AI (GenAI) models—primarily GANs, VAEs, GPTs, and diffusion models—can autonomously generate urban data, scenarios, 3D city models, and designs within urban digital twins (UDTs). It synthesizes evidence across transportation, energy, water, buildings, and urban planning to propose a GenAI-enabled vision for cognitive digital twins and outlines technical strategies for integrating GenAI into various software architectures. Key contributions include a taxonomy of GenAI methods, a sector-by-sector mapping of autonomous data and scenario generation capabilities, and a forward-looking framework for human–AI partnership, privacy-preserving data, and scalable 3D city modeling. The study highlights opportunities and challenges in data quality, privacy, ethics, governance, and computation, offering a roadmap for deploying GenAI-enhanced urban digital twins to support reliable, scalable, and participatory smart city development.
Abstract
The digital transformation of modern cities by integrating advanced information, communication, and computing technologies has marked the epoch of data-driven smart city applications for efficient and sustainable urban management. Despite their effectiveness, these applications often rely on massive amounts of high-dimensional and multi-domain data for monitoring and characterizing different urban sub-systems, presenting challenges in application areas that are limited by data quality and availability, as well as costly efforts for generating urban scenarios and design alternatives. As an emerging research area in deep learning, Generative Artificial Intelligence (AI) models have demonstrated their unique values in data and code generation. This survey paper aims to explore the innovative integration of generative AI techniques and urban digital twins to address challenges in the realm of smart cities in various urban sectors, such as transportation and mobility management, energy system operations, building and infrastructure management, and urban design. The survey starts with the introduction of popular generative AI models with their application areas, followed by a structured review of the existing urban science applications that leverage the autonomous capability of the generative AI techniques to facilitate (a) data augmentation for promoting urban monitoring and predictive analytics, (b) synthetic data and scenario generation, (c) automated 3D city modeling, and (d) generative urban design and optimization. Based on the review, this survey discusses potential opportunities and technical strategies that integrate generative AI models into the next-generation urban digital twins for more reliable, scalable, and automated management of smart cities.