Citiverse: Virtual Worlds for Urban Systems
Header-Image Credit: Optics: crystals exhibiting interference colours by R.H. Digeon, ca. 1883.
What is the Citiverse?
Over the past few years, several technology streams that had been developing independently have started to converge around a shared question: What would it mean to have a city-scale virtual environment where people can experience, test and negotiate urban futures together?
The word “Citiverse” describes this convergence. It was coined in the context of the ITU Global Initiative on Virtual Worlds and AI, launched in 2024 by the International Telecommunication Union, the United Nations International Computing Centre and Digital Dubai. The initiative defines a citiverse as a network of interconnected virtual worlds that represent and are synchronized with their physical urban counterparts, using AI, digital twins, extended reality and real-time data to address the needs of cities and their inhabitants.
What makes the concept more than a rebranding of “Smart City 2.0” is the convergence itself. While mainstream planning practice often relied on static master plans, urban theorists have actually understood cities as complex, interconnected systems for over a century—the challenge has always been finding the tools to operationalize this theory.1 Visual technology has matured from cinema to consumer-grade VR and AR. Connectivity now supports real-time synchronisation of large, multi-user environments. And AI has reached a point where it can populate virtual models with agents, generate scenarios and translate between data types that previously did not speak to each other. Individually, each of those are a tool. Together, they promise to open up the possibility of shared environments where a planner could walk through a proposed redesign, a citizen could experience what a new tram line means for their street and an energy engineer could overlay heat demand data on the same model and all at the same time.
What makes the concept more than a rebranding of “Smart City 2.0” is the convergence itself. Urban theory has spent a century moving from static master planning toward understanding cities as complex adaptive systems. Visual technology has matured from cinema to consumer-grade VR and AR. Connectivity now supports real-time synchronisation of large, multi-user environments. And AI has reached a point where it can populate virtual models with agents, generate scenarios and translate between data types that previously did not speak to each other.
Individually, each of those are a tool. Together, they promise to open up the possibility of shared environments where a planner could walk through a proposed redesign, a citizen could experience what a new tram line means for their street and an energy engineer could overlay heat demand data on the same model and all at the same time.
Components of this vision are already deployed in cities like Seoul2, Helsinki3 and Zurich4 for participatory planning, risk management and infrastructure operations. But a fully integrated, city-scale Citiverse remains an ambition. What exists today is a fast-moving landscape of pilot projects, policy frameworks and international standardisation efforts laying the groundwork. On the policy side, the EU established the CitiVERSE EDIC5 in February 2024 with 14 member states and over €80 million invested and the first UN Citiverse Challenge launched in 2025.
All of this suggests that the question is no longer whether cities will build virtual counterparts, but how, for whom and under what governance conditions.
The Citiverse as convergence point of different technology and knowledge streams (slide snapshot).
Current Projects and Initiatives
I have been working on these topics from several angles at Fraunhofer IAO within the Urban Systems Engineering department.
ITU and international standardization: I contributed to the Citiverse Use Case Identification Track, one of the working groups within the ITU Global Initiative. The track developed a taxonomy of 50 urban use cases across five thematic areas, each scored for SDG alignment, scalability, impact and feasibility, and mapped along a maturity horizon from deployed to experimental. The full taxonomy and thematic reports were published in mid-2025. Fraunhofer IAO’s Steffen Braun is among the contributors. My involvement focused on the urban planning and placemaking dimensions and on connecting the taxonomy work to our applied research.
Metaverse Themenwochen at Fraunhofer IAO: I am the departmental face for the campaign “Metaverse: Vielfalt erleben, Innovation gestalten”, which brings together researchers from across Fraunhofer IAO to demonstrate how the convergence of digital twins, AI, XR and connectivity creates practical possibilities for industry, cities and people. I wrote the opening blog post for the campaign and chaired the Open Lab evening at the Zentrum für Virtuelles Engineering in March 2026, where visitors could experience VR demonstrations in our CAVE6, interact with the telepresence robot Reachy, try AI-driven co-creation tools with Hyperfy, and explore neurocognitive research setups. I have also presented on these themes in a talk titled “LListening to the City: Approaches in Bridging Complexity and Clarity”, which explored how AI foundation models might help translate between quantitative urban data and qualitative lived experience. See: Poster PDF.
The Urban Futures Teleporter: One idea that emerged from this work is a concept for a mobile, immersive VR/XR environment housed in a recycled shipping container. The Urban Futures Teleporter would allow stakeholders to virtually travel through time at a specific location in their city: To experience how their street looked in the past, how it looks today, and how it might look under different climate and development scenarios. It would build on the Morgenstadt network’s City Lab methodology and integrate multisensory elements to make abstract climate data tangible. This remains a vision, not a funded project. You can read about the concept here: Urban Futures Teleporter Vision (PDF).
Background: Why this interests me
My interest in the Citiverse is not primarily technological. It comes from a problem I have encountered repeatedly in my work on EU urban projects like URBREATH and SPARCS: The difficulty of creating shared understanding across stakeholder groups that speak different professional languages, care about different things and operate under different constraints.
The deeper I look at this, the more I think the underlying issue is about the relationship between the tangible and the intangible in how we understand and govern cities.
The tangible side is what gets measured: GDP, traffic flow, housing costs, energy consumption. These indicators are the foundation of decision-making because they are comparable, aggregable and fit well into dashboards. For most parts they are site-inspecific and consist of SI-based or monetary units.
The intangible side is what makes a city actually liveable: Community cohesion, perceived safety, cultural identity, the quality of public space, the atmosphere of a neighbourhood at night. These things depend on small details on the ground. They are non-linear, context-specific and resistant to standardisation. The process of quantifying them often strips away exactly what made them meaningful in the first place.
Neither side alone is enough. The tangible can tell you how many people use a park but not whether the park feels safe. The intangible side may tell you if a neighborhood feels attractive but cannot directly translate these results to cost-effective maintenance schemes. Governing a city well means holding both together.
Here’s the catch: The higher up you move in a planning or management hierarchy, the harder that becomes. The more things you need to take into account, the greater the pressure to simplify, aggregate and compare, and the greater the reliance on indicators that can turn complex reality into a single number. This creates a structural pull toward the tangible, not because anyone thinks it is sufficient, but because it is what works at scale.
The field’s response has been to try to close the gap from the indicator side: Developing ever more nuanced metrics that attempt to make the intangible measurable, like e.g. liveability indices, social cohesion scores, walkability metrics, perceived safety surveys and hundreds more. My colleague Aapo Huovila documented over 1,500 smart and sustainable city indicators in his dissertation. That number is not a sign of failure; it reflects a genuine effort and ambition to give value and leverage to the urban intangibles. But it raises a question I keep coming back to: Does this proliferation follow a pattern?

Open Question: Does a cycle like the following exist? …a complex urban question arises that existing indicators cannot adequately capture. The available standardised metrics e.g. from official UN or EU indicator frameworks, are too broad or generic for the local context. Local practitioners develop new, context-specific indicators to fill the gap, adding to a growing total. Over time, governing bodies seek consolidation and produce new standardised lists, which in turn are again too broad for local context. The cycle restarts. Each iteration does produce useful governance tools and insights, but the overall pattern raises a question: Whether the gap between what indicators capture and what stakeholders need to understand is one that more indicators alone can close. Whether this loop exists systematically is an open question. The pattern of proliferation itself is observable.7
The indicators themselves are valuable, they have materially improved how cities set targets, allocate budgets and track progress. But adding more of them does not resolve the underlying problem: That quantitative data and qualitative experience are different kinds of knowledge, and no amount of the former automatically produces the latter.
What the Citiverse idea suggests, at least to me, is the possibility of complementing this cycle rather than just repeating it. Not by replacing, consolidating or helping choose the right indicators, but by adding a different kind of interface between data and understanding. Not a dashboard with more rows but a shared environment where quantitative models and qualitative experience can coexist. Where instead of reading a table about urban heat, you sense or see the temperature difference between a street with tree cover and one without. Where instead of debating abstract revenue and cost figures, you stand at the spot and see what a tree would do for the space. The data is still underneath, but it becomes accessible through experience rather than expertise alone.
I do not want to overstate where we are. Most of what I have described is either common-sense, early-stage, experimental or still conceptual. The technology is advancing very fast. Questions about governance, data privacy, equity of access and the risk of immersive manipulation are largely unanswered. And there is a real danger that the Citiverse becomes just another layer of complexity rather than a bridge through it.
But the underlying question feels right to me: How can we build tools that help people across different backgrounds arrive at a shared, experience-based understanding of their city, one that honours both the things we can count and the things that count but cannot easily be counted?
That is what I am trying to work on.
Worth reading
A few recent publications that I found particularly interesting:
-
Zheng et al. (2025): “Urban planning in the era of large language models.” Nature Computational Science. Tsinghua and MIT Senseable City Lab argue that LLMs can fundamentally change how we plan cities, from spatial cognition to simulated community engagement. DOI: 10.1038/s43588-025-00846-1
-
Xu et al. (2025): “Using human mobility data to quantify experienced urban inequalities.” Nature Human Behaviour. Shifts the measurement of urban inequality from where people live to where they actually go. A interesting example of what becomes visible beyond conventional indicators. DOI: 10.1038/s41562-024-02079-0
-
Zhang et al. (2024): “Urban Foundation Models: A Survey.” KDD 2024. The first systematic definition of Urban Foundation Models. Essential for understanding the technical landscape underneath the Citiverse idea and how language, vision, time-series and multimodal urban data converge in a single model architecture. DOI: 10.1145/3637528.3671453
-
Stojanovic et al. (2026): “Blind spots and actionable insights for urban governance of the climate-biodiversity-health nexus.” npj Urban Sustainability. Documents the problem this page discusses: How sectoral silos, fragmented targets and soft governance tools create blind spots in urban policy, even when the data is technically available. DOI: 10.1038/s42949-026-00345-w
-
Sanchez et al. (2024): “The Ethical Concerns of Artificial Intelligence in Urban Planning.” Journal of the American Planning Association. AI ethics in one of the top planning journals and a counterweight to the enthusiasm: What happens when these tools meet real governance, real bias, real power asymmetries?
-
Coyne, R. (2025): AI and Language in the Urban Context: Conversational Artificial Intelligence in Cities. Routledge. Related to the “LListening to the City” idea: Coyne explores what happens when we treat cities not just as systems to be optimised but as communicative environments and what role conversational AI plays in there. DOI: 10.4324/9781003535751
Related links
- ITU Global Initiative on Virtual Worlds and AI
- Citiverse Use Case Taxonomy Overview (ITU Publication)
- ITU Citiverse Thematic Reports
- EU LDT CitiVERSE EDIC (European Commission)
- 1st UN Citiverse Challenge
- Fraunhofer IAO Metaverse Campaign
- Blog: “Das Metaverse als Schnittstelle zwischen Modell und Wirklichkeit”
Footnotes
-
Viewing the city as a complex, adaptive system isn’t a new idea, even if it took a while for everyday planning practices to catch up. Early thinkers like Patrick Geddes (Cities in Evolution, 1915) and Lewis Mumford (The Culture of Cities, 1938) were already using biological metaphors to describe urban interconnectedness. By the 1960s, Jane Jacobs famously called cities problems of “organized complexity” (The Death and Life of Great American Cities, 1961). The journey since then has largely been about developing the right tools to actually map and manage this complexity, from early computer models (like Jay W. Forrester’s Urban Dynamics, 1969) to network sciences (Michael Batty, The New Science of Cities, 2013).. with the newest devlopments in Urban Foundation Models and the Citiverse … ? ↩
-
Seoul: Metaverse Seoul for virtual municipal services; S-Map digital twin for environmental and urban simulation. ↩
-
Helsinki: Open-data Helsinki 3D+ platform for citizens and planners to virtually experience new urban developments. ↩
-
Zurich: Digitaler Zwilling Zürich merges 3D geospatial and BIM data to simulate climate scenarios like urban heat mitigation. ↩
-
EDIC = European Digital Infrastructure Consortium headquarter in Valencia. ↩
-
CAVE = Cave Automatic Virtual Environment, a room-sized immersive VR setup where 3D visuals are projected onto walls and floor. ↩
-
Slide from a seminar presentation. See full Presentation. ↩