1 Introduction

Urban Digital Twins (UDTs) have emerged with the promise of becoming transformative tools for urban planning and management. As cities globally adopt them, approaches diverge in use cases, governance, openness and their relation with stakeholders (D’Hauwers et al., 2021). UDTs hold the potential to support overcoming urban challenges and complexity by bridging the siloed structures of cities, which are reflected in the data management, city departments, and involved stakeholders (Weil et al., 2023).

However, many of the current UDT implementations appear as stand-alone applications tackling explicit urban challenges, but are not interoperable with each other. Therefore, attaining the full potential requires an ecosystemic approach that integrates stakeholders, technologies, and data across interconnected systems (Bennett et al., 2023).

Scholars have emphasized the dual role of cities as both physical and social entities (e.g. Batty, 2024; Bettencourt 2024). For UDTs to surpass the capabilities of existing city models and provide meaningful advantages across the multitude of urban planning and governance tasks, a deeper integration of physical and social systems is essential. This integration necessitates a modular, participatory and challenge-driven approach to their implementation (Nochta et al., 2021). Moreover, unlocking the real potential of UDTs requires a shift of the planning of digital twins from systems built for individual and isolated use cases to networked, ecosystemic frameworks.

By designing digital twin applications with interoperability in mind from the early stages, they can be seamlessly connected to form a broader digital twin ecosystem. This approach enhances their value and usefulness, allowing digital twins to move beyond their current role as isolated tools (Bennett et al., 2023).

In this work, we present a development agenda for UDTs, building on our experiences from the City of Helsinki and FinEst Centre for Smart Cities. We base this development agenda on two assumptions: Firstly, we assume that an ecosystemic approach could effectively support UDTs in navigating the dynamic and heterogeneous technological landscape of the cities. Secondly, from the technical perspective, we argue that interoperability of data and tools will be a central enabling factor in accomplishing this. We firstly review the current state of UDTs, focusing on Helsinki and FinEst Centre for Smart Cities (Sect. 2.), after which we present the key aspects for further development of the UDTs, consisting of establishing common understanding (Sect. 3.1), modularity & interoperability (Sect. 3.2), presentation agnosticism (Sect. 3.3) and the social dimension (Sect. 3.4).

2 Current State of Urban Digital Twins

Since the emergence of the term “Digital Twin”, it has been widely adopted in various disciplines, with diverging definitions ranging from static digital models to autonomous cyber-physical systems. The application of digital twins has gained significant traction throughout the last decade in the context of cities and the built environment (Abdelrahman et al., 2025).

For digital twins of cities or urban environments, a variety of terms are used. These include: urban digital twin, digital twin city (Deng et al., 2021), city scale digital twin (Nochta et al., 2021) smart city digital twin (Francisco et al., 2020) and local digital twin (Raes et al., 2025; Villanueva-Merino et al., 2024). Today, there is no consensus on the definition of what exactly an (urban) digital twin is, and many authors have highlighted the plurality of the concept (Abdelrahman et al., 2025; Deren et al., 2021).

Digital Twins of the urban environment, in this publication “Urban Digital Twins” (UDTs), are perceived as virtual replicas of cities that combine data from multiple sources to create a comprehensive digital representation of urban environments. They hold potential to serve as powerful tools for city planners, policymakers, and stakeholders to monitor, simulate, analyze, and optimize urban systems before implementing changes in the real world (Azadi et al., 2025). Specific digital twin applications support urban asset management and participatory practices. These digital representations hold the potential to revolutionize planning and governance by enabling more up-to-date analysis and future projections based on data-driven insights.

Understandably, UDTs have, and are, being researched and implemented globally with varying foci. Stemming from urban planning, UDTs have initially emerged based on static 2D maps and 3D city models. From this static foundation, the digital representations gradually were enhanced with additional information, e.g. semantic information and dynamic data (e.g. Schrotter & Hürzeler, 2020; Lehner & Dorffner, 2020).

While UDTs already integrate previously siloed data within a single application, most remain constrained to specific use-cases, relying on limited input data, models and simulations. This use-case driven approach represents an intermediate stage in the progression towards a comprehensive digital representation of urban environments. Advancing beyond this stage is crucial for developing a holistic representation of urban systems, supporting the development of smarter, more efficient, and sustainable cities.

2.1 Urban Digital Twins in Helsinki

In Helsinki, the development of UDTs has been strongly built upon prior work with 3D city models. Currently, the city maintains two central city model assets: a textured mesh model (updated by a renewed city wide survey) and a CityGML-based model (updated along with the base map). Both of these models cover the entire administrative area and are openly available (Helsinki, 2025). To further integrate these models to various urban planning and management processes, a development project has been launched at the Helsinki Urban Environment Division. The work began with the development of proof-of-concept level prototypes of both tools and processes that utilize the 3D city models and other geospatial data. The completed projects have been documented as videos, with their results guiding further development work in the city administration (Helsinki Urban Environment Division, 2025).

Forum Virium Helsinki, the city’s innovation company, has focused on the development of new application areas and data sources of UDTs (Virtanen et al., 2024), further developing the concept of socio-technical UDT (Ruohomäki et al., 2024) and studying the potential for forming a digital twin of urban mobility (Forum Virium, 2024). Currently, several projects are ongoing related to facilitation of data ecosystems (DataLiiKe, 2025), data marketplaces (SEDIMARK, 2025) and data spaces (TFDS, 2025). These activities contribute to the further integration of data ecosystems, data spaces and UDTs.

2.2 Digital Twins at FinEst Centre for Smart Cities

FinEst Centre for Smart Cities is an international and transdisciplinary research and development centre, operating as an independent organization under Tallinn University of Technology (TalTech). Established in 2019 by TalTech, Aalto University, Forum Virium Helsinki and the Estonian Ministry of Economic Affairs and Communications, the centre focuses on enhancing the quality of life in urban areas. Following this aim, the Centre is researching Urban Digital Twins globally and developing (prototype) digital twin applications within local projects and in collaboration with researchers from various international research institutes, citizens, city officials and leaders (Soe, 2017; FinEst Centre Homepage, 2025).

FinEst Centre for Smart Cities has led the GreenTwins project, focussing on two topics: First, the development of prototypes for a dynamic vegetation layers for the Digital Twins of the cities Tallinn, Estonia and Helsinki, Finland. This layer includes algorithmically generated 3D models of plants, along with data on their growth patterns and seasonal variations under local climatic conditions. Second, GreenTwins brought up two digital twin applications named Virtual Green Planner and Urban Tempo, designed to engage urban stakeholders, especially citizens, in the design of urban green areas (Prilenska et al., 2023; FinEst Centre, 2023).

Research beyond the integration of biotic layers to UDTs, FinEst Centre for Smart Cities led the development of the Renovation Strategy Tool (ReSTO). ReSTO is a digital-twin based platform for municipal decision-makers enabling the optimization of required investments into their building stock, by assessing alternative scenarios across various aspects, including technology, building and infrastructure design, urban planning, and costs. By leveraging digital twin data from the Estonian Building Registry (https://livekluster.ehr.ee/ui/ehr/v1) and other public databases, ReSTO allows to determine economically optimal renovation strategies while considering predefined constraints such as budget availability, energy performance targets, and environmental objectives (Arumägi et al., 2023; FinEst Centre, 2025).

In addition to leading its own research projects, the FinEst Centre for Smart Cities, actively participates in multiple international projects developing UDTs, such as the urbanLIFEcircles project. Here, the Centre contributes to the development of digital twin applications for urban biodiversity monitoring, management and modelling (LIFE Public Database, 2025).

3 Key Elements of Future Urban Digital Twins

3.1 Establish Common Understanding and Standards

As already mentioned (Sect. 2.), a concise and universally applied definition of an UDT does not exist yet. Currently, the inconsistent use of the term “urban digital twin” for various differing concepts hampers the advancement of both the discourse and development in the field. This lack of conceptual clarity also risks reducing “digital twin” to another short-living buzzword, rather than a meaningful enduring framework for urban innovation.

Thus, the establishment of a common understanding, clear definition and common standards for UDTs is necessary, to support both discussion and the development itself. In addition, the UDTs should be conceptually positioned in respect to a number of surrounding concepts, such as:

  • 3D City Models (Biljecki et al., 2015)

  • Urban Dashboards (Kitchin et al., 2016)

  • Urban Data Platforms (Soe et al., 2022)

  • Metaverse/Cityverse (Kshetri et al., 2024)

The question on which different authors tend to have diverging opinions is whether a digital twin necessitates automatic feedback into the real world or not. While this requirement originates from a fairly agreed-upon definition of a digital twin in manufacturing technology, it has proven to be less applicable to UDTs. It has been noted that many of the UDTs don’t meet the criteria of a digital twin, if a strict definition is applied (Metcalfe et al., 2024).

For our purposes, the inclusion of automated actuators bridging the virtual and the digital world are seen as one additional component in the digital twin framework. However, as the development of UDTs is continuously evolving, this distinction may become increasingly relevant in defining technology readiness and technology maturity models in the future.

Standards and information models related to UDTs can be approached from a holistic perspective, aiming for a model covering all of the object types with their semantics, or with a minimalist approach, ensuring only the object identifiers (Ellul et al., 2024). Multiple ongoing attempts to unify the definition of UDTs can be identified, including at least the following ones (Table 1.). While reaching a cross-sectorally accepted definition might not prove to be feasible now, having an internally coherent understanding of what is meant by a UDT would greatly benefit cities by reducing confusion and supporting internal collaboration.

Table 1. Examples of UDT definition work.
Full size table

3.2 Modularity and Interoperability

An ecosystemic UDT leverages the principles of interoperability and modularity to create a collaborative environment where various stakeholders can contribute and benefit from the system (Bennett et al., 2023). This requires at least a certain degree of common understanding (Sect. 3.1). The approach aligns with the concept of data spaces, which provide a shared digital infrastructure for data exchange and collaboration (Gil et al., 2024). Thus, realizing functional data ecosystems is a relevant task for UDT development as well.

Without a true ecosystem of data, applications and tools, UDTs risk remaining as cities’ internal urban planning tools. As the problems encountered in urban settings do not follow administrative boundaries, interoperability of tools is also needed to allow collaboration between administrative units and beyond.

To ensure data protection, privacy and compliance with relevant regulations (such as GDPR), robust security measures and ensuring data sovereignty is crucial for maintaining trust among stakeholders and protecting sensitive urban data.

Realizing UDTs as a networked, modular system (Fig. 1.) involving multiple stakeholders requires the adoption of standardized data formats: Adopting common data standards and formats ensures seamless integration and interoperability between diverse data sources and different components of the UDT. Thus, modularity and interoperability are connected. Additionally, an API-driven architecture enables maintaining a clear master data repository and provision of updates by the responsible entity/organisation. Implementing a robust API layer allows for seamless communication between different modules and external systems, and integration of real-time data and new components such as AI agents or simulation engines. This way, different UDTs may also have their respective use cases or foci.

Fig. 1.
Diagram illustrating a flow chart with five hierarchical levels: "Users," "Use-cases," "Tools," "APIs," and "Underlying systems." Icons represent different entities at each level, such as a graduation cap, building, factory, and group of people for "Users," and a tree and car for "Use-cases." Arrows indicate interactions and data flow between levels, connecting various elements like cubes for "Tools," circles for "APIs," and cylinders for "Underlying systems." Dotted and solid lines show different types of connections.figure 1

UDTs formed as an ecosystem of data, APIs and tools enabling data flow across different use cases and users, with returning data fed back to underlying systems.

Full size image

3.3 Presentation Agnosticism

Digital Twins are primarily presented (and viewed) through various visualizations, including interactive 3D environments, renderings of 3D models and VR/AR. However, their presentation may also incorporate audio feedback, data-driven analytics, and interactive simulations to provide comprehensive and dynamic representations of physical assets or systems (Mrosla et al., 2025).

As UDTs are expected to support decision-making processes, it is essential that they remain agnostic to their means of presentation. This means that the underlying data and models should be independent of any specific presentation tool or technique.

By separating the data and analysis layers from the presentation layer, UDTs can:

  1. 1.

    Accommodate diverse user needs and preferences for (data) presentation.

  2. 2.

    Enable the integration of new presentation technologies as they emerge.

  3. 3.

    Support multiple simultaneous presentations of the same data for different purposes or stakeholders.

  4. 4.

    Support the integration of different external data sources in the presentation medium suited for them.

While the data models commonly used in UDTs, such as CityGML for 3D city models, support both 3D visualization and other types of thematic visualizations (e.g. semantic or analytical representations; Fig. 2.), the implementations of other than geospatial visualization are extremely rare in the UDT context.

Fig. 2.
Aerial view of an urban area with a series of bar charts and pie charts on the right. The bar charts display data on building completion years and heights, with the tallest building reaching 54.53 meters. The pie charts illustrate building facade materials and purposes of use, highlighting categories such as business and residential. Total buildings count is 848. The visual provides an overview of urban development and architectural characteristics.figure 2

An example of CityGML building models being visualized in 3D (left) and as a statistical plot of their properties (right), illustrating how same data can be visualized in different tools to serve different use-cases. Data and 3D visualization courtesy of the City of Helsinki.

Full size image

3.4 Social Dimension

As UDTs are in the end intended to serve citizens, even if by non-direct influence e.g. by providing better decision making environments for the administration, it is worth questioning how well the UDTs are able to cover the complexities of life in cities. Multiple authors argue that current digital twins are merely abstractions of reality, omitting many elements (Batty, 2018). Furthermore, while digital twins are often developed with the intention to enhance public participation e.g. in decision-making processes, this promise remains largely unfulfilled (Charitonidou, 2022).

To address this gap, the “social dimension” in UDTs development and application should be further strengthened, bridging the gap between social processes, social phenomena, urban population and the urban digital twin (Ruohomäki et al., 2024). Following Ruohomäki et al (2024), the social dimension of the UDTs is here understood to include:

  • Socio-economic and demographic data

  • Inclusion of non-physical spatial artifacts such as legal and administrative boundaries, and management units

  • Data from participatory actions and crowdsourcing

  • Inclusion of social phenomena and artifacts

The importance of the human dimension is further highlighted by most of the envisioned UDT applications being “human in the loop” systems, supporting decision making in urban planning, management and other operations with improved insights, simulation results and tools (Schrotter & Hürzeler, 2020; Lehner & Dorffner, 2020). While more autonomous systems have been explored in the smart city context, they are typically present in cases related to interpreting sensor data streams from smart infrastructure (Mohammadi & Al-Fuqaha 2018). For processes like urban planning, little is known about how to practically include ethics and other non-tangible values into the potentially autonomous decision making systems–at the same time the role of social processes and artifacts has to be acknowledged for these systems to be truly feasible (Charitodinou, 2022).

4 Discussion and Conclusions

The conception and development of Urban Digital Twins is happening simultaneously at many fronts:

  • The technologies involved in digital twins evolve. These include sensor/IoT systems, algorithms, models (and associated standards) and presentation methods such as through VR/AR. Thus, continuous adaptation is needed for the UDTs to leverage the most suitable available and emerging tools.

  • Urban planning, management and participatory processes continue to integrate upcoming technologies and data, leading to new and improved use-cases for the UDTs.

  • The conceptual discourse on UDTs progresses, potentially leading to a more consensual understanding of what the UDTs actually are.

For this evolving system to be functional, usable and effective, the development of digital twins of cities should not occur haphazardly, depending on available funding and driven by hype-cycles. Instead, proceeding in a well-orchestrated approach, where the intertwined components and their relationships to each other are well thought out, allows forming a functioning, updatable and useful ecosystem of tools, data and technology. Effectively leveraging this in practice can lead to a comprehensive change of how cities are planned and managed: from siloed domain-specific approaches to a holistic understanding of a city.

In this work, we have identified four key elements for future development of UDTs, them being 1) common understanding, 2) modularity & interoperability, 3) presentation agnosticism and 4) the social dimension. While these aims may never be fully agreed upon or reached by all UDT stakeholders, we argue that even a limited progress towards them would be highly beneficial for cities.