What is a digital twin? A virtual clone of a real system, such as a factory, a building or a city, which uses data to mirror its current state and behaviour. A digital twin is not a static model, but a dynamic one that changes along with its physical counterpart, thanks to real-time data streams. It sounds like a great tool for engineers, but it is much more. Cities all over the world are discovering that local digital twins (LDT) mirroring their systems are a powerful tool to foster innovation and citizen engagement.

Northern Europe hosts some of the most stimulating examples. Several cities, such as Amsterdam, Stockholm and Zurich, have created interactive 3D LDTs of their urban areas to simulate the impact of new construction developments, let contractors test new designs digitally, and collect feedback from developers and residents. Helsinki the capital of Finland, has been working with digital twin technology for more than two decades. At first, the city used 3D interactive models to evaluate architectural competitions for new developments. Now, Helsinki’s urban planning digital twin has a much bigger goal: it has become an Energy and Climate Atlas, modelling data on energy consumption at the building level, to help achieve carbon neutrality by 2030. In Finland, heating buildings accounts for up to 56% of the city’s carbon emissions. The digital twin’s powerful engine can measure solar radiation and shadows on each surface of each building for every hour and day of the year (which is important in Finland's northern latitudes where the sun angle changes drastically during the seasons). The digital twin’s high level of accuracy (20 centimetres, compared to Google Maps’ 5 meters) is used to help developers and owners calculate the most effective mix of heating systems. In other words, with this tool Helsinki's homeowners can compare the cost of different heating options and their environmental impact.

Digital twins are not only useful for planning and testing scenarios, but also for creating new opportunities and services. For instance, Helsinki's LDT was the basis for a competition among companies to find the best ways to use open data, 5G networks and augmented reality. A very advanced digital twin model for mobility and transportation is in the works (by the LiiDi2 project) to combine public and private data sources, design new services and open new business opportunities.

The value added of these models is that they can link different data streams from city systems that are usually managed separately, such as mobility and air quality. Belgium's Flanders region, through the DUET project, is working on a mobility model that that can predict the traffic impact of closing streets and the effect on the air quality and the environment. The port of Rotterdam in the Netherlands is creating an ambitious LDT as a 3D city operations platform, with multiple objectives such as visualising energy consumption at the building level, monitoring mobility and improving public safety. For example, the LDT tracks both road and waterway traffic intensity through a broad network of cameras and sensors. It uses this information to optimise the opening and closing of bridges to avoid congestion, while providing guidance to boats and vehicles. 

Tampere is another Finnish city having invested in multiple LDTs over the last decade, also through the STARDUST project. One of the LDTs is a digital twin of the Hervanta area, where automated driving vehicles have been tested. Another LDT is a 3D model of residential buildings, which is used to measure and optimise energy efficiency and consumption, with the aim of reducing CO2 emissions. Rodrigo Coloma, from the city spatial data unit, says: “We are always collecting data and information to create 3D models for infrastructure and urban planning. The new construction law will allow in 2025 the use of 3D-models to approve petitions in the process of construction permits”.

Tampere also launched a “data-driven city for citizens” programme, which uses data to understand citizens’ needs and to design new services. Anniina Autero, the head of the data-driven city flagship Safe Pedestrian City Project, says: “We have a strong technology ecosystem in Tampere, but we want to put people at the center centre of our work, use technology for the people, not the other way around”. Tampere is collecting data and creating proof-of-concept digital twins for four new pedestrian services, ranging from improving pedestrian experience to developing cultural walks. Autero explains: “We define specific indicators about the quality of experience, such as safety (which areas are safe to walk in?), family needs, and climate impacts. We monitor them with sensors that feed into the data model and experiment with different combinations that can improve people’s satisfaction”. The process also involves more creative methods: for example, Tampere used the Minecraft education tool to let children play with the 3D city model and design their own environment. “In the future, we would like to connect the virtual and the physical environments to a metaverse concept – a cityverse” concludes Autero.

Helsinki and Tampere are among the forerunners, but the number of European cities engaged in LDTs is surprising. According to a recent study by the European Commission there are 135 LDT initiatives in 25 EU countries, but they vary in their level of sophistication. About half of them are still in the initial phase, developing 3D data models and planning for future developments. Another 57 initiatives are more advanced, integrating multiple domains and enabling simulations and scenarios based on historical data. Only 12 platforms are considered “predictive twins”, which can use real-time data to simulate and test different outcomes. The use of AI techniques is still not widespread. The main applications of LDTs are in urban planning, urban mobility and environmental management, aiming to reduce CO2 emissions.

According to Sophie Meszaros, project manager of the Open and Agile Smart Cities initiatives (OASC), cities and communities  are just starting to fully understand the the multiple potential applications of local digital twins.  “The technology is ready,” she says “but the initial investments are high and the cost-benefit analysis is not always clear”. A large part of the initial investment consists in collecting, making available and interoperable datasets from different sources, or installing networks of sensors to provide real-time data flows. Meszaros points out the challenge of trust: decision makers without specific technical knowledge may not anticipate all the benefits of the exploitation of data, or trust the results of a simulation.  “The key is the availability, quality and reliability of the underlying data,” she says “this is why we are investing time and effort in dataspaces for smart cities”. Meszaros is leading the initiative ‘European Data Space for Smart and Sustainable Cities and Communities’ (DS4SSCC) which fosters data sharing and develops methods to improve data quality and availability, in order to build trust. 

Several other initiatives by the European smart cities community seem promising for the future take-up of digital twins. The Living-in.eu community has created a specific forum on LDTs, where city, regional and national experts share use cases and interests from different groups across the EU. The Eurocities network has a specific Task Force on Digital Twins. The European Commission (EC) has launched funding calls to develop a Local Digital Twins toolbox, which will offer cities re-usable tools, reference architectures, open standards, and technical specifications. Finally, in June 2023 the EC launched a funding instrument to upscale Digital Twins towards the Citiverse , through a European Digital Infrastructure Consortium (EDIC) led by Estonia and supported by Germany, Slovenia, Czech Republic and Spain, open to all other EU countries.

By Gabriella Cattaneo

Mapping of Local Digital Twins (LDT) in Europe by Use Case and Maturity

Legend: LDT levels of maturity: Awareness = 3d data model, political commitment; Experimental = multidomain integration, simulation only on historical data; Predictive = allows simulation and what-if scenarios with real-time data, automated decisions enabled by humans.

Source: European Commission, Directorate-General for Communications Networks, Content and Technology, Robalo Correia, A., Sousa, M., Mulquin, M. et al., Mapping EU-based LDT providers and users, European Commission, 2023, https://data.europa.eu/doi/10.2759/547098