Periodic Reporting for period 4 - UMnD (Urban modelling in higher dimensions: embedding generalisation of 3D data in a 4D model)
Reporting period: 2021-03-01 to 2022-11-30
In this project, we developed a fundamental solution for providing 3D data at application specific LoDs. For this, we investigated an innovative method beyond the state-of-the-art by developing a 4D data structure to store the multi-LoD knowledge as an extra dimension to the 3D spatial dimensions in an integrated 3D+LoD (4D) model.
We have also developed novel reconstruction methods for 3D urban data at application specific LoDs. This provides the LoDs for the 4D model. Finally, we developed a methodology to compare two existing 3D city models, associate and store them with their correspondences in 4D, as well as a methodology for versioning to extract and update the model using one LoD at a time (as a slice of the 4D model). This opens a new horizon for modelling parametrisable aspects of urban data in one data structure, like temporal aspects. More research is needed to implement the 4D approach in practice. This requires addressing the complexity added to the process by the set of rules needed to enforce the management of the model.
The use case driven project results for the automated reconstruction of application specific 3D+LoD data from different sources and the publication of these countrywide 3D data as open data in our developer friendly-format has made 3D data accessible to many more urban professionals as well as to the wider public. This has significantly increased the use of 3D data in a wide variety of urban applications, such as simulations of energy demand, wind flow, pollutant dispersion, solar energy potential, noise, and the urban heat island effect. By making 3D building models accessible and affordable, they can finally become mainstream and be used in simulations and applications to optimally design, maintain and change urban areas to address global challenges.
We have studied how to reconstruct 3D data at different application-specific LoDs as well as how to generalise 3D data at multi-LoDs from one 3D dataset. This provides the Level of Details for the 4D model. In use cases with urban specialists, we defined both the generic and application-specific needs for 3D+LoD data and developed our data-solutions accordingly, which made it possible to advance in fields that are important to address urban challenges (noise, wind flows, energy, urban heat stress etc).
Major achievement in our project is the methodology to automatically reconstruct detailed 3D models of buildings for large areas, together with their lower LoDs. We have applied the methodology to the whole of the Netherlands resulting in a nationwide Digital Twin in the CityJSON standard that was also developed by our team (which is another major result of the project, adopted as community standard by the worldwide standardisation body OGC). The resulting multi-LoD 3D data that we published as open data are frequently downloaded for use in a wide variety of urban applications and has been adopted by several governmental organisations. Our 3D building data service has opened the use of 3D data beyond the traditional urban data-users making the wide use of 3D urban data in other research fields affordable and possible. This nationwide 3D multi-LoD data set has also become the main data source for our own developments and experiments.
In addition, our GeoBIM research has yielded novel results for the integration of highly detailed Bulling Information Models (BIM) in lesser detailed 3D geographical contexts. We also developed an innovative method to describe urban morphology in 3D. Finally, a versioning methodology has been developed to store the versioning knowledge of 3D data in one data structure which can represent versions at different LoDs as well as versions at different points in time. Our versioning approach enables to manage the evolution of multi-LoD datasets, such as the efficient tracking of changes for individual LoDs and the ability to extract and update the model using one LoD at a time, as a slice of the 4D model.
We have succeeded in developing the fundamentals for this approach (i.e. 3D generalisation and reconstruction methods to obtain 3D data at different LoDs; 3D+LoD (4D) data structure; 4D visualiser). We have also studied new principles that support the success of the proposed methodology as well as 3D urban data applications in general. More specifically, we have developed a method to validate and clean 3D data, we have developed a standard to compactly encode and manipulate 3D urban data, and we have studied and identified the 3D+LoD data needs of a wide variety of urban applications. Our studies on use cases for 3D+LoD data for urban applications and our solutions for automatically reconstructing application-specific 3D+LoD data for large areas accordingly increased the use of 3D urban data in a wide variety of urban applications. This is new to the prevailing method of collecting and maintaining one 3D dataset that had the almost impossible objective of serving needs of any application and that required significant postprocessing for specific use cases.
Our studies on the integration of 3D city models and Building Information Models (BIM) of individual buildings, have yielded new opportunities for the integration of highly detailed BIM models in lesser detailed 3D city models.
We also developed a method to elevate building metrics that describe the urban morphology, currently often limited to 2D, into full 3D which is fundamental in a broad range of investigations across the fields of city planning, transportation, climate (heat stress), energy, and urban data science.