The application of Digital Twin (DT) technology is rapidly increasing, and Dutch planning practitioners are facing several challenges. The translation of legal texts into 3D efect simulations is not without validation issues. General rulings of legal texts require local interpretation. And the 'fairness' of representations of new urban designs and policy efect measurements are part of the struggle. This paper explores six DT cases and presents the findings and some areas for bridging the gap between traditional law and the DT. Based on these cases, a synthesis consisting of nine validation challenges is formulated and future research directions are presented.
Dutch governments are using Digital Twin (DT) Solutions at an increasing rate to support
spatial planning discussions among stakeholders, operational asset management, and permit
processing for citizens. DT technology has become one of the major development programmes
on the Digital Agenda of both regional and city innovation platforms [
The research question is framed as: ‘Which validation challenges do governmental
organizations in the Netherlands encounter while aiming for a fair translation of policies, laws and
regulations into Digital Twin solutions for climate governance?’ We match those validation
challenges against the existing set of FAIR principles, being 1) Findability, 2) Accessibility,
3) Interoperability, and 4) Reusability [
In the context of spatial planning, Schrotter & Hürzeler [
For our purpose, we use the definition of a DT that is being used for spatial planning where
a digital copy of a specific existing physical environment is made that represents the legal
boundary, characteristics, and regulatory conditions of that area in one integral and consistent
representation. This is a juridical extension of the definition used by Schrotter [
In the Netherlands, the legal environmental spatial planning platform Ruimtelijkeplannen
[
The legal tradition of environmental administration was based on extensive PDF files with
long deliberations. The influence of GIS knowledge on the legal profession was also noticeable
at an early stage since ‘map-people’ created the legal term ‘search area’, which at the time did
not exist as a concept in the legal profession [
The research presented here is explorative in the sense that we seek to set contours for the problem space and at the same time assist the government agencies involved to explore the fairness and validity of the DT, which to our knowledge has not been done before. To provide a structured overview of the challenges of legal (regulatory) DTs we have chosen to investigate six cases where province and city management seek to build a legally consistent DT for diferent reasons. The number six is chosen because this set proved to be rich enough to raise challenges that require further research. The approach is that of action research, the researchers are involved in supporting cities and provincial councils, and consulting on standardisation validation internally on a professional basis. The identification of validation challenges was carried out in an unstructured way during design discussions with the responsible design teams, the cases were centrally discussed during these sessions. In addition: eleven webinars and online team meetings were organised in 2020 and 2021 by Dutch umbrella organisations VNG [15] and IPO [16] discussing the development of DTs among software suppliers and civil servants of many Dutch cities and provinces. Another two sessions in 2020 were dedicated to the semantic harmonisation of the terminology used in legally binding spatial plans. Specific research was carried out in the context of the Healthy Urban Living programme [17] and the Feed eParticipation EU project [18]. In addition, 22 interviews (eleven semi-structured and eleven open interviews) were held specifically with the professionals engaged in the building of the digital infrastructure for the National environmental act (DSO) [19]. Based on the data collection, a multitude of challenges were identified. The strength of this action research using multiple explorative cases is also its weakness. The validation challenges were gathered over time under diferent circumstances in discussion with government peers, subject-matter experts, domain experts and legal experts. Formalisation of the method using structured interview protocols would often have disturbed the richness of the discussion. For the analysis of the data, three filters were put in place to achieve a level of objectivity: 1) The challenges listed were mentioned at least three times on separate occasions, 2) The challenges listed were mentioned by at least three diferent respondents, and 3) The challenges had to applicable to at least three domains, for example, the built environment, emission, and mobility as depicted in a DT.
The province of Utrecht and the City of Amersfoort together worked on the mapping of the characteristics of a specific area around the railway station into a DT to support the decisionmaking process of the redesign of that area. The main goal of the DT was the ability to visualize in 3D the diferent design aspects and discuss these in co-design with the stakeholders of the railway station neighbourhood. In particular, the need to manage the (relatively new for the Dutch) rain flooding problem and the heat stress problem caused by too much concrete and too little green. Diferent planning scenarios and estimated cause-and-efect relations were mapped into the DT. The legal environment concerns city planning rules. The strategic policy level involves many and sometimes contradicting climate-related policy issues, such as solar energy ifelds, less car-based mobility and clean sewer systems and avoiding air pollution.
The Environmental act digital platform 2022 is a major programme to advance the Dutch spatial planning process toward more integration, fewer pre-set rules and better (digital) government service for the citizen. It combines 22 sets of regulations concerning environmental planning into one integrated workflow. The ‘applicable rules’ are business rule translations of juridical texts and they are called applicable because they should enable the citizen to click on any object in the Netherlands and retrieve all governing rules related to that object. These applicable rules should answer the permit request of a stakeholder to act employing webforms. This is enacted by the design of ‘smart’ forms following a dynamic decision tree as a permission check beforehand. The set of business rules is feeding the requested requirements for entering a specific permit per theme or activity type.
This DT has been designed in the context of the Environment Act (as of the first of July 2022), The DT is used to visualize national datasets, such as the real estate or building data, road infrastructure data and ‘green’ areas and subsurface data. In addition, the city of Utrecht applies the DT for participation with stakeholders. This DT is based on Unity’s game engine. The advantage of this DT approach is that one can, unlike with more traditional GIS platforms, enlist the development power of the gaming industry to develop the 3D city model of the city of Utrecht. The development of this DT is a collaboration between the cities of Amsterdam and Utrecht. The aim is that all Dutch municipalities should be able to use modules built based on this open platform.
The Royal Institute of environmental control and disease control (RIVM), The Utrecht University and the province of Utrecht collaborate on a project called ‘the Green Benefits Planner’ to discuss scenarios with a positive efect on climate policy goals with stakeholders. The project is part of a larger regional initiative called the Healthy Urban Living programme [17]. The DT is part of the scenario builder and depicts the efects of diferent design decisions such as planting trees, creating water channels or new cycling areas and themes such as heat stress. The goal of the Green Benefits Planner is not to show one thematic map, but to balance many themes. The balancing was implemented in the software called Tygron. To provide policymakers with a simple control panel that shows policy goals, the ‘policy room for manoeuvring’ and the efect of those choices on the living environment.
The Markerwadden, as a design decision, was a multi-million Euro compensating action for building in a Natura2000 (thus heavily guarded by law) protected area in the Dutch grand sweetwater lake. The Markerwadden is an artificial island for ‘green’ and biodiversity purposes. This was a result of an extended negotiation period. The next step was to reach an agreement about the exact spatial design of the protected area. The DT was created because of the many conflicting interests. The goal of this juridical DT was to enable stakeholders to consider other stakeholders’ interests in the planning process. In this case, the Natura2000 was conflicting with the water directive, so a clear-cut legal solution for several land allocation purposes could not exist.
This case originated in the process of designing green electrical transportation over the water towards Dutch city centres. This DT was developed to support the Provincial council debate about the role of the province in the emerging digital landscape. The province and the cities with more and more shipping of (internet) goods to the city centre, were pushing for climate goals, but the market required intervention in cheaper fossil-based transport over land. The DT depicts the terminal area and the dilemmas that require government intervention, such as deploying a blockchain platform for identity management. The DT was built in the game platform Unity, together with suppliers of the infrastructure who provided DT equivalents of their products in a reusable object library.
To examine the impact of the identified validation challenges we map them against the FAIR
principles [
1. Findability and visualization: versus the level of detail. The challenge is to ofer the right level of detail for the right purpose. In some cases, this is a risk due to ‘expectation management’ where showing too much detail at the wrong time or in front of another target group raises sentiments of the finality of the concept where the message meant was much more open for discussion between scenarios. The public stakeholders also tend to focus on details which were not meant to be part of the scenario target theme by the experts. The level of detail among professionals and in the situation of a court case tends to be much higher than what is politically necessary during scenario debates. 2. Consistency: The reduction of large descriptive texts into binary schemes, i.e., digital ‘yes’ allowed’ versus ‘no, not allowed deontic logics. The standardization of policies into binary Yes or No answers for permission requests is required for eficient service delivery to citizens in a digital front ofice of the city. 3. Legal semantics and accessibility: the mapping of legal terminology (semantically) is often too diverse to harmonize suficiently for a binary purpose. The policymakers are eager to preserve room for interpretation and city spatial plans tend to have a very high level of freedom without much standardization, which contrasts with the service delivery aim of that same city. There are also semantic issues concerning juridical terms, especially those with a juridical history. Juridical terminology is precise to exclude non-intended exemptions. The demands on digital service delivery through smart forms and business rules require accessible language for the end-user. Successfully mapping the ‘folksonomy’ to the legal taxonomy (applicable rule) is a matter of interpretation and it makes it harder to maintain consistency. 4. Interpretation and consistency: In the Netherlands, a culture of case-based law processing exists that is not compatible with the perspective of a consistent nationwide infrastructure; diferences in the notion of case handling as a paper-based administrative process versus a standardized spatial data infrastructure that has to work on a country level scale. 5. Legal: authorization of modelling of policy criteria for design planning sessions and monitoring. For example, the water flooding models, heat stress models and mobility models as described earlier represent the problem of validation and the ‘democratic mandate’ of the supplier of those models. In one region there were three heat stress models in use by diferent agencies and developed by three diferent suppliers. This leads to confusion about which is the right or ‘oficially recognized’ model. Government agencies are seeking a new role in certifying the models in use [20]. 6. Integrality and interoperability: Specialist sectoral applications do need to become interoperable and exchangeable for the sake of better collaboration and re-use of data. The challenge in a DT is to create inter-sectoral validation. 7. Consistency: Correct (transparent, logical, explainable) translation or mapping of legal texts into object languages and business rules. This includes correct representation of the cause- and efect relations concerning the variables in (informal) design sessions versus those that will stand in more formal (e.g., court or tendering) situations. This includes issues of timing and expectation management, especially in participative trajectories. It also includes ‘applicable rules” and forms during the permit stage. 8. Archiveability: Historical consistency for court cases (and ownership of the historical database). The legal consistency and legitimacy entail the existence of legal memory. The higher administrative courts require ‘time stamping’ or the ability to travel back in time and obtain the ‘picture’ of the case many years back. The challenge is to generate that integral picture over time as a consistent whole, including contextual information. Although GIS systems and DTs can log many variables and data layers, it is hard to determine what a ‘complete’ contextual picture should look like, especially given the increasing technological capabilities over time. 9. Interoperability: Supplier modelling interoperability and modular software development.
Separating business rules from their representation by DT vendors (separation of concerns). The business rules have to be explainable so that public values can be audited, especially if they have a legal consequence [21, 22]. If this is not done, public values cannot be guaranteed in the public stack due to the lack of transparency.
Our research at this point confirms that the mapping of policies, laws, and regulations into object attributes for DTs should receive more attention in light of the FAIR principles. This was true for single business rules that power webforms for citizens and it is even more important in DTs with more integrated climate themes. The ‘validation gap’ between the abstract Fair principles of Findability, Accessibility Interoperability, and Reusability and the validation challenges as found in this research should be bridged by a maturity model that fits with the object-oriented policy cycle. The logic is that climate goals programmes in the end seek to address attribute values of objects in the physical environment and governments work in object-oriented policy cycles to achieve them. This cycle is speeding up due to increasing demands on scarce space and conflicts with climate goals. The Dutch Courts of Law are taking an increasingly firm position in this debate. The need to build correct representations of the spatial environment in all its aspects and themes (e.g., sound, water, biodiversity, economic activity, mobility, pollution) is therefore increasing, [23, 24]. For both the overall environmental strategy design and during implementation, government agencies need to be increasingly consistent in the mapping between design plans, project plans, and the existing legal-administrative landscape [25, 26, 27]. Citizens would feel ‘cheated’ if the DT scenarios were unfeasible in the juridical reality of constraints and boundary values, such as policy goals on climate. There is a complex mapping process in play here that pose potential risks for the valid implementation of the law and regulation. The models that feed a DT depict the actual space for manoeuvring while balancing policy goals for the relevant authorities. Integrality is another reoccurring theme. The DT infrastructure enables measuring variables of one theme, e.g., planting trees aimed at emission reduction against another theme, e.g., the mobility efect of less roads. The data in a DT represents a wider range of themes, data sets, standards, criteria, semantics, and assumptions about cause- and efect relationships to enable scenario design. The visualization of the spatial environment in DT is thereby shifting from static digitalized spatial plans with links to the relevant legal texts in large PDF’s towards ‘dynamic DTs’. These DTs depict de potential efect of diferent legal regimes on climate related aspects of the environment in a number of design scenarios [28]. The thematic integrality as described above and the regional level of detail would require DT solutions that go beyond the 3D representations of cities, especially for the legal interoperability.
The proposed variables that should be measured and scored between the policy cycle phases at all DT layers are: 1) Integrality: are policy themes taken into account in all phases? 2) Consistency: are datasets, definitions, values and norms consistent in all phases? and 3) Object status: are object attributes and mutations (of roads, buildings, biodiversity, air quality, etc.) in the environment monitored in all phases?
One of the questions is whether the research question is valid outside of the spatial planning
domain and climate issues. Other authors, such as Schrotter [
Journal on Advances in Systems and Measurements 9 (2016) 210–219. [12] K. Smit, M. Zoet, An organizational capability and resource based perspective on business rules management, in: Proceedings of the International Conference On Information Systems (ICIS, 2018, p. 1–17. [13] A. Fuller, Z. Fan, C. Day, B. C, Digital twin: Enabling technologies, Challenges and Open
Research. IEEE Access 8 (2020) 952–971. [14] K. Yong-Woon, Digital twin maturity model. web 3d 2020 industrial use cases workshop on digital twin visualization, 2020. [15] V. N. Gemeenten, Vng, 2022-03-02. URL: https://Vng.Nl/. [16] Rijksoverheid, Informatiepunt leefomgeving, 2022. URL: https://Iplo.Nl/. [17] Gezondstedelijklevenhub, Gezond stedelijk leven data en kennishub, 2022. [18] R. Peters, The law, the map and the citizen: Designing a legal service infrastructure where rules make sense again, 2016. [19] R. Peters, K. Smit, J. Versendaal, Responsible ai and power: Investigating the system level bureaucrat in the legal planning process, 2021. [20] Hiemstra, Vries, Wat provincies kunnen doen in een digitaliserende wereld, 2022. [21] M. Lokin, Wendbaar wetgeven - de wetgever als systeembeheerder, 2018. [22] R. Doesburg, T. Engers, Guidelines for nmas applied to calculemus-flint, JusletterIT (2020). URL: https://doi.org/10.38023/107c571f-cc75-4c8b-9cff-40dced0f2283. doi:10. 38023/107c571f- cc75- 4c8b- 9cff- 40dced0f2283. [23] G. Blair, Digital twins of the natural environment, Patterns 2 (2021). [24] J. Hetherington, M. West, The pathway towards an information management framework-a ‘commons’ for digital built britain, 2020. [25] L. Sutton, C. Beardall, J. Lam, L. Standen, What are digital twins and what are the legal issues with them?, 2021. URL: https://Www.Gtlaw.Com.Au/Knowledge/ What-Are-Digital-Twins-What-Are-Legal-Issues-Them. [26] Y. Hofs, Rechter verklaart stikstof gedoogbeleid provincies onwettig, 2022. [27] T. Engers, Goed geregeld? het recht als ontwerpvraagstuk, 2004. [28] S. Knapp, V. Coors, The use of eparticipation systems in public participation: the veps example, Urban and Regional Data Management 105–116 (2007).