By Dr Adam Mowlam, Manager Smart Cities, City of Greater Geelong

In the ever-evolving landscape of urban development, local governments are increasingly embracing innovative technologies to streamline their operations and enhance decision-making processes. One such technology gaining prominence is three-dimensional building models. Here, we take a closer look at the concept of building models, their varying levels of fidelity and their significance for local government professionals.

In the context of spatial digital twins, a building model is a highly detailed digital representation of a physical structure within a virtual environment. Designed with precision to accurately emulate real-world buildings, these models offer local government professionals an in-depth understanding of the urban landscape.

Building models display various levels of information, often referred to as Levels of Development (LOD). Typically, these levels range from LOD0 to LOD3, each indicating a unique level of intricacy and detail. The LODs can be categorised as follows:

LOD0: This level corresponds to a basic building footprint without attached height information or visible building parts. It is suitable for 2D cartography and analysis.

LOD1: This level includes height information, even allowing buildings to consist of several parts, each with a different height. The result may appear blocky, it strikes a good balance between visual appeal, data capture cost and performance. This mode is preferred for tasks like vehicle navigation and large area analysis.

LOD1 building models

LOD2: This level introduces roof shapes and detailed architectural elements of the building facade. LOD2 is suitable for 3D visualisation of larger city models and their surrounding context, although it’s common to use LOD1 in suburban areas and LOD2 across activity centres.

LOD2 textured building models

LOD3: This level introduces into the building model detailed wall and roof structures, balconies, bays and projections. Generally considered cost prohibitive at large scale, it’s common to mix buildings with LOD3 level information within a LOD2 dataset.

Understanding these different levels of fidelity allows local government professionals to choose the appropriate building model for their specific use case, balancing accuracy with resource requirements and considering the intended use of the digital twin.

Building models vs 3d photomesh

While building models and 3D photomesh may appear similar initially, they serve distinct purposes within the digital twin domain.

Typically, building models represent physical structures, crafted from geometric data. Conversely, 3D photomesh relies on aerial or satellite imagery to construct a textured, three-dimensional model of the terrain, encompassing both the built and natural environment. The primary divergence lies in the depth of detail and accuracy. Building models excel in facilitating in-depth analysis and simulation, allowing local government professionals to evaluate the impact of alterations on individual structures.

On the contrary, 3D photomesh is better suited for visual representation and contextual understanding, providing a comprehensive view of the entire urban landscape without the intricate details inherent in building models.

Exploring sightlines in 3D building models.

Why building models are useful for local government

The integration of building models into digital twins offer a range of benefits to local government professionals.

Building models facilitate precise spatial analysis allowing the evaluation of proposed changes on individual structures and the broader urban environment. The building models are critical in comprehending how new developments or infrastructure projects might impact the existing landscape, extending to signal propagation and line of sight analysis for enhanced telecommunications planning.

Leveraging building models, local governments gain the ability to visualise and scrutinise zoning regulations, ensuring alignment with city planning standards (e.g. height controls, building setbacks). Building models often play a pivotal role in emergency response planning due to the detailed information of the urban environment. This ranges from assessing the impact of natural disasters, identifying critical infrastructure locations, and evaluating the repercussions of emergencies on specific buildings.

Infrastructure development projects benefit from valuable insights such as simulating the integration of new structures, assessing their impact on existing infrastructure, and optimising layouts for maximum efficiency.

Building models also serve as compelling visual aids for proposed developments, facilitating effective communication with the public, enhancing transparency and community engagement. Residents can easily visualise and comprehend potential changes in their neighbourhood, fostering a sense of inclusivity in the planning process.

Shadow analysis with 3D building models.

Products available to local government

Local governments have access to a diverse array of 3D building products, encompassing both commercial and authoritative options.

Commercial Products

Google and Apple Maps: Highly used and robust platforms, both products rely on AI to generate building models that are combined with aerial/satellite imagery, street-level views and a range of other two dimensional data to offer a comprehensive perspective of the urban landscape. Outside of the CBD the building models are generally LOD1.

NearMap, OpenStreetMap and Here: Numerous commercial products employ high-resolution aerial imagery and AI capabilities, contributing to detailed and up-to-date building models. These products are more easily accessible within corporate GIS systems and often provided as a suite (trees, road infrastructure, swimming pools).

Geoscape Buildings: Geoscape is another authoritative product integrating AI technology, it delivers high-quality geospatial data, including detailed 3D building models, offering intricate information about building footprints, heights, and roof structures. The linkages between the dataset with Cadastre, Property, and Australia’s Geocoded National Address File (G-NAF) make it highly usable.

Authoritative Products

VicMap 3D Building: Recognised as an authoritative source, VicMap 3D Building, enriched with AI capabilities, stands as a pivotal building data repository for all Victorians. Developed by the Victorian Government, it ensures accurate and up-to-date information regarding building footprints, heights, and other pertinent details. The availability of these products varies across Australia.

Using building models for smart city development in geelong

The Geelong Smart City journey has always centred on the extensive use of data for informed decision-making. To advance this commitment, a crucial step involves a comprehensive process of acquiring data and modelling buildings.

Data acquisition: capturing foundations for precision

City-scale high-fidelity building models hinge on the availability of 3D data products, typically taking two primary forms. Aerial imagery involves high-resolution photos, serving as foundational data for subsequent modelling and textures, while LiDAR scanning generates detailed 3D point clouds, providing precise elevation data required for building heights and rooflines.

In the Geelong project, two pivotal datasets were employed. Firstly, a spatially accurate 3D photomesh was crafted from 7.5cm vertical and oblique imagery spanning a 180 square km area.

Additionally, a geo-referenced high-density LiDAR with a swath accuracy <5cm was utilised to ensure optimal feature extraction capability. These datasets underwent conversion to Cesium 3D Tiles and were aligned to GDA2020 with ellipsoidal height to enhance visualisation capabilities.

Building modelling: from footprints to digital realism

Historically, the building modelling phase would begin with building footprint extraction (LOD0) from aerial imagery using automated algorithms. This process helps to best determine scope and scale of the project area, in this case Vicmap 3D

Buildings was used. Once completed, the data capture process saw the transformation and reprojection of the complete 3D tileset, migrating it from Cesium Ion to glTF files. This step enables access to all tile levels in Blender, the software used for 3D modelling/editing, facilitating the creation of high-quality LOD2 buildings. The facade detailing varies across buildings through the addition of facades and includes the removal of trees and cars, including from textures.

Integration and continuous evolution in digital twins

Following the creation of the building, models are integrated into the broader digital twin platforms such as Digital Twin Victoria. This is a key step as it involves aligning the created models with geospatial data layers (such as planning zones) to really capture the power of viewing the built environment in context of cartographic layers.

The City’s strategy involves providing access to this data for City staff, property developers/planners, investors, the community, and other essential stakeholders. The Geelong Data Exchange will serve as the platform for hosting this data. Given the substantial city-wide investment of over $17 billion in capital projects, ensuring the currency of data is crucial. We are currently in the process of implementing a robust procedure to address this need.

Local governments can significantly enhance decision-making in urban development by investing in the development of building models.

These models, when integrated into digital twins, offer urban professionals valuable insights for precise spatial analysis, refined urban planning, and strategic infrastructure development.

The Geelong case study exemplifies the benefits of building models in optimising zoning regulations, fostering public engagement effectively, and providing crucial support for the city’s future growth and development. By harnessing the transformative power of building models, local governments ensure transparent, resilient, and community centric decision-making, ultimately shaping cities to be more efficient, sustainable, and well-prepared for the future.

Related articles

Leave a reply

©2024 Infrastructure Magazine. All rights reserved


We're not around right now. But you can send us an email and we'll get back to you, asap.


Log in with your credentials

Forgot your details?