Digital twins are fuelling the development of new business models that promise better, more efficient service offerings. And while we've seen some successes in areas like manufacturing and design, it is not clear whether XaaS-driven approaches have a good fit with, and can deliver the same benefits when applied to the digital twins of smart cities. Determined to investigate that, we developed a typology of local digital twins to serve as a basis for future examination of various commercialisation strategies for different LDT types.
An ongoing trend towards more flexible consumption is changing the nature of offerings that companies bring to market. Today, we see more and more technology providers offering anything-as-a-service, or XaaS, solutions to meet evolving customer needs. XaaS companies leverage cloud computing and the Internet of Things to offer customers the ease of access without the responsibility of ownership and maintenance. With XaaS, there is no need to build everything in-house. Organisations can just purchase a XaaS license to take advantage of its extensive infrastructure. What’s more, instead of paying huge sums of money for a complex package, clients can sign up only for those functionalities and services that they really need. And because XaaS and IoT are often bundled in a single offering, users benefit from consolidated access to real-time data across multiple solutions. This way they can manage every data point in the network, whether it’s a supply chain, public transport infrastructure or street lighting system.
Digital twins combine IoT, cloud computing and other advanced technologies such as Artificial Intelligence to provide a clone of a physical entity which can then be monitored, interrogated, remote-controlled, simulated under different conditions and even left to manage a physical counterpart with little or no human supervision (an AI controlled digital twin). As the popularity of digital twins grows within and across sectors, so does the interest in technology’s ability to drive business model innovation, prompting many potential adopters to wonder whether digital twins can offer innovative business models with completely new value propositions.
To answer that we only need to look at recent manufacturing innovations that embraced XaaS. One example would be a digital twin operating under the Factory-as-a-Service model. The ‘smart factory’ concept allows users to purchase manufacturing and service hours instead of investing large amounts of capital in equipment. Unburdened by high investment costs and risks, manufacturers can focus on things that matter most (e.g. client satisfaction) by adapting their processes to more flexible production. Crucially, this model makes small-batch manufacturing not only possible but commercially viable, as the digital twin can easily produce customised products in any quantity required to accommodate changing demand. SmartFactory as a Service can already be seen in action today in Munich’s dynamic factory quarter.
Similarly, an asset oriented digital twin operating under the Product-as-a-Service model allows users to purchase a desired outcome rather than equipment. For instance, rather than buying a robot to weld two pieces of metal, a company might instead purchase several wielding operations ad hoc. So when it comes to PaaS, products are offered via subscription along with services. In this model, what-if studies, predictive maintenance, on-demand performance analysis and simulations can be sold as a one-time service or be included in a recurring service contract. Because asset ownership is not transferred to the customer, PaaS has a huge sustainability potential. It pushes the owner to deliver a better product which can be reused, repaired, recycled and redistributed, and not simply to sell as many products as possible.
From smart lighting to water pumps, the number of PaaS use cases is on the rise. Perhaps the best known example is Rolls-Royce’s TotalCare. It removes the burden of engine maintenance from the customer and transfers the management of associated risks to Rolls-Royce. Under this model, the company is responsible for proactively managing the engine through its lifecycle to achieve maximum flying availability. In doing so, it reduces waste and optimises resource efficiency, which explains why Rolls-Royce calls TotalCare a circular business model.
What TotalCare and other XaaS examples show is that interoperable digital twins open up a whole range of opportunities for data-driven business models. This pioneering technology creates an important shift in value creation, which now comes less from a physical asset than from a digital twin used to service a product and optimise every step of the value chain.
The Smart City Context
For DUET, the main challenge with existing digital twin business models is that they largely apply to use cases in manufacturing and engineering. There is a notable lack of relevant frameworks for other sectors, including smart cities. In an attempt to fill this gap, we set out to develop a typology of local digital twins (LDTs) and their exploitation scenarios. The term LDT was recently coined by the European Commission to describe digital twins of cities, their processes, systems and assets. They operate at a larger scale but use similar advanced technologies as private-sector counterparts (e.g. cloud computing, IoT, AI) to improve the way cities are planned and managed. The term LDT is synonymous with the concept of urban digital twins, as both refer to digital twins at urban scales.
LDTs offer many of the same advantages as other smart city systems. For instance, with LDTs you can get contextual information on physical assets (e.g. buildings’ energy consumption). You can measure the pulse of a city (e.g. mobility patterns) and other key trends impacting the city and its inhabitants (e.g. pollution levels, climate change). You can even optimise city’s performance by making its systems run more efficiently; an example would be the optimisation of a transport network through an AI powered control center. But what makes LDTs different compared to IoT dashboards, GIS, BIM and other tools is that they interconnect various urban data sources and modelling algorithms in a way that can grow with the city as it becomes more complex. Through this dynamic bidirectional communication link, LDTs mirror the vibrant, complex and evolving nature of its physical urban counterpart, which makes them an ideal sandbox for policy experimentation.
When viewed from this angle, LDTs can provide a risk-free testing environment for simulating alternative policies to improve city management. They can reveal policies that underperform and identify leverage points for interventions that may succeed. This ex-ante evaluation function makes LDTs a coveted tool among public authorities. And while city administrations are often the main users of LDTs, there are plenty of international examples demonstrating LDT use beyond policy making. A digital twin that is open to the ecosystem can benefit other actors, too. For instance, it can help citizens experience public services better and faster, while companies can develop new products and services based on LDT data.
Given these benefits, it's not surprising that interest in the LDT concept is growing. Over the years, we have witnessed more than a dozen LDTs emerging in different parts of Europe, from Paris to Stockholm, from Herrenberg to New Castle. Some of these are still in the prototype phase, but many are already operational. In the future, we can expect the LDT community to expand significantly thanks to Living-In.EU, Gaia-X, EIF4SCC and other initiatives that aim to create an enabling environment for LDTs to become mainstream.
As the LDT landscape becomes more crowded, there is a need for a theoretical framework that is broad enough to capture a variety of LDT use cases, both existing and future ones. DUET’s typology is the first step in this direction, and it’s part of a bigger research on LDT business models that we intend to complete in 2022.
Input for the initial stage came from knowledge accumulated during the project, as well as recently conducted primary research that involved interviews with representatives of several cities that already have a LDT or are in the process of creating one, among them Vienna, Helsinki, Rotterdam and Örebro.
The Typology of LDTs
When we think of business models, we tend to think of ways in which a company does business to make a financial return. A firm-level business model typically describes the value offering to a customer group and how the value is delivered to generate sustainable and profitable revenue streams. It’s much harder to conceptualise the business model at the level of smart city, where value generating activities transcend the boundaries of individual actors and even sectors. Here, a multitude of interactions taking place vertically (from organisation all the way to the ecosystem) and horizontally (e.g. B2B) enables collaborative innovation to emerge and shape the creation of public value. A collective business model that explains how value is being co-created among different actors belonging to a network therefore lends itself to the study of smart cities. And it is for this reason that we used a network-centric approach as a prism through which to examine LDTs.
The application of a network-centric approach to the study of LDTs leads to several exploitation scenarios that differ according to usage and control. The former is mapped along the y-axis to denote primary LDT users and beneficiaries. The usage parameter moves from a closed model (with government being the sole user) towards an open one in which a digital twin is open to all: civil servants, residents, companies. In the latter case, benefits from LDT usage accrue to different stakeholders, generating an external value for the ecosystem.
As smart cities are essentially large data ecosystems, we mapped control parameters along the x-axis to denote owners of data resources and technical infrastructure. Control can be centralised (regulated by the government) and decentralised (regulated by the ecosystem). In a decentralised environment, different actors serve as data sources. They are in control of information they hold and can set conditions for data sharing and access.
This mapping resulted in a four-quadrant typology for examining LDTs, as shown in the diagram.
The Typology of LDTs. Source: Authors' elaboration based on D2.4
The Closed LDT
The Closed LDT represents the most centralised LDT type in our typology. It is intended as an internal decision support tool for local administration and so is not open to the ecosystem. In the Closed LDT, all data comes from government sources. Closed LDTs are initiated, managed and paid for by the government as the derived value directly benefits the city. The government controls conditions for data access and sharing. It also determines what components to use for digital twin development. Closed LDTs, although designed for internal use, are usually built using open-source software as cities generally prefer not to purchase proprietary solutions to avoid vendor lock-in.
An example of the Closed LDT is Vienna’s GeoTWIN, which is currently in the prototype phase. It has been created to achieve several objectives. One is to break down data silos within government by linking different sources to the city information model e.g. GIS, census data, socioeconomic data, energy consumption data, maintenance data. Another is to aid urban planning and the development of new infrastructure projects, such as the North-West Train station area. In the case of Viena's geoTwin, the value is created internally within government in the form of improved policy making, reduced risk and enhanced urban planning, among others. The costs are covered by the GeoTWIN department using government funds.
Digital geoTwin of the City of Vienna. Source: Lehner & Kordasch 2020
The Network Oriented LDT
The Network Oriented LDT is a step up from the Closed LDT. In this scenario, the city initiates a digital twin project with the intention to make it available to the wider community. The cost of implementing the Network Oriented LDT is typically borne by the government. This covers the technical infrastructure, as well as the gathering and structuring of data.
Given its outward orientation, the Network Oriented LDT supports decision making in the public and private sectors. As the digital twin is no longer confined to government use, the city and businesses alike can create new services using insights and data provided by the platform. Thus the Network Oriented LDT leverages public funding to open the innovation space to the ecosystem.
The Helsinki Digital Twin is an example of the Network Oriented LDT. It was created by the city to spur collaboration across multiple city functions and provide a collective response to urban needs, including how to use energy resources more efficiently. In fact, one of the main applications of Helsinki’s digital twin is the Energy and Climate Atlas, which is an open web service built on a semantic city model.
First published in 2018, the Atlas includes basic information on all buildings and their energy consumption. The 3D model of the city allows users to make comparative analysis to find out e.g. how much energy different buildings consume, how much solar energy can be received by different buildings. The Atlas is a useful information tool for housing companies, property developers, manufacturers of solar panels, and also citizens. The Atlas can be used by anyone, even for commercial purposes as long as the the City of Helsinki is credited as a data source.
The Helsinki Digital Twin. Source: https://kartta.hel.fi/help/internet/fi/3d.html
The Network Owned LDT
Sitting in the top-right corner of our typology is the Network Owned LDT. It is here that we see the highest level of diffusion with regards to usage and control. Whereas other LDT types are driven by cities, the Network Owned LDT is the initiative of an ecosystem, which subsequently oversees its implementation and management, leaving the government to play an important facilitating role along with industry, science, research and civil society. The Network Owned LDT is not an exclusively city-funded project. Here, the costs are shared proportionally by actors that signed up to the initiative.
Needless to say, true examples of the Network Owned LDT are hard to come by in real life, at least for now. City administrations are still the main driving force behind most urban digital twins. (Projects like DUET are an exception, but because our use cases are managed by partners representing local and regional governments, the statement about the driving force still holds true.) They are the initial innovator due to investment requirements and upfront costs. They pay for data collection and management, and decide what technology to use for digital twin development. Thus, at present, the Network Owned LDT is more an aspiration than reality. It is the next evolutionary state in the LDT maturity model that could be achieved in the future if the right conditions are in place.
So what are the requirements for the Network Owned LDT? For this type of urban digital twin to emerge, we would need an open digital ecosystem where information and services from public sector, private sector, academia and citizens can be made available in a trustworthy manner. Essentially, this would have to be a federated data ecosystem of networked spaces based on agreed rules to promote interoperability. This in turn would enable data and service providers to collaborate with each other and to link, develop and further expand their offerings more easily. The good news is that efforts to develop such ecosystem are already underway. EU-level legislative developments (e.g. the Data Governance Act) and the hitherto mentioned initiatives like Gaia-X and EIF4SCC leave one sanguine about the prospect of the Network Owned LDT becoming a reality sooner than one might expect.
In fact, some early signs of transition toward the Network Owned LDT can be discerned in existing urban digital twin projects. A case in point is Rotterdam. The city wants to set up a digital infrastructure to bring together local actors with a view to encouraging them to share information and services within the data ecosystem. Currently, there are some 350 different smart city initiatives that are not connected to each other. The platform would change that by interconnecting as many Rotterdam-related applications as possible through the use of open standards. As well as breaking down the silos, the final solution will help build a shared picture of reality while offering more opportunities for all (civil servants, companies, citizens) to participate in the city’s ecosystem.
Rotterdam 3D. Source: https://www.3drotterdam.nl
The Network Enabled LDT
The Network Enabled LDT also exhibits high levels of decentralization when it comes to data and infrastructure, however its objectives are mostly policy related which makes the government a sole user of the solution. Although data comes from the ecosystem and some infrastructure is shared with the local administration, access to the actual tool and results is restricted to policy makers. The Network Enabled LDT thus represents a closed part of an otherwise open ecosystem where government can set up their own digital twin instance for private use.
As with the previous type, a true Network Enabled LDT has yet to emerge in real life. Nevertheless, we noticed some manifestations of this urban twin in Örebro. The city is moving towards a linked data ecosystem (“Ett sammankopplat Örebro”) to facilitate the exchange of information on social housing, urban development, education, transport and mobility, among others. The local administration plans to combine some of this data with its own sources to improve the functioning of the city. For instance, one of the goals is to leverage sensor data, some of which is owned by private companies, to enhance road network’s performance. The expectation is that the IoT based traffic signalling system will help reduce congestion thanks to more accurate information on traffic density.
A Summary of LDT Characteristics
Our initial research has shown that existing LDTs differ according to levels of control and usage. The least open types are the Closed LDT and the Network Enabled LDT. Being policy oriented, their main purpose is to support decision making within local government. These digital twins are not available to the ecosystem although sometimes it provides data to supplement governmental sources.
The most open types are the Network Owned LDT and the Network Oriented LDT. These digital twins are accessible to the ecosystem and so create opportunities for actors other than policy makers to benefit from LDT data, for example by creating new services (companies) or becoming more engaged in democratic life (citizens).
Today, pretty much all LDTs are government-led projects, which leaves true ecosystem owned or enabled examples in the realm of future possibilities. One might argue that this status quo will continue into the future. For us, however, the question about whether the two decentralised types will emerge is not a matter of if but when. The reason for our cautious confidence: several EU initiatives (Gaia-X, DGA etc.) that aim to establish open, interoperable cross-domain digital ecosystems in which data and services can be shared between providers in a trusted and secure manner to create new value propositions.
Our typology identified four LDT scenarios but it stops short of prescribing which one is best for what city. The choice clearly depends on local needs and priorities, on data that has to be included and the desired maturity level of a digital twin. For some cities, the Network Owned LDT will be a natural target determined by recent progress in data governance. For those that simply want to enhance urban planning, the Closed LDT may be a more appropriate option.
It’s also worth noting that LDTs are not mutually exclusive. A local authority may decide to have one for internal purposes and another one with different functionalities for the wider public. Importantly, various LDTs can coexist, reinforcing each other and perhaps acting as parts of a bigger digital twin linking other cities in the same country and even across borders.
To test its robustness and usefulness, we will be mapping more cities onto our framework in the coming months. In parallel, we are going to start the second phase of our research on LDT business models, where the main focus will be on exploring the commercialisation potential of different LDT types. As part of this, we will be investigating whether LDT stakeholders can benefit from cloud business models like XaaS in the same way as stakeholders in manufacturing do. If you have an opinion on that, let us know in the comments below.
Authors: Ruben D'Hauwers (imec-SMIT-VUB), Pavel Kogut (21c Consultancy)
Smart HR Tool.