A digital twin is a virtual model of a real building, district or city that is kept in step with the real thing through data. For a real estate developer or a giga-project, that ranges from an explorable 3D model used to sell and plan, up to a live operational model fed by sensors and building systems. What separates a twin from a 3D render is synchronisation: a twin changes when the asset changes.
This guide defines the term, sets out five levels of fidelity so you can say which one you are buying, and lists what it takes to build and keep one running.
What is a digital twin?
The most widely used definition comes from the Digital Twin Consortium, which in December 2020 defined a digital twin as “a virtual representation of real-world entities and processes, synchronized at a specified frequency and fidelity”. Two words in it do the work. Frequency is how often the model is updated, from once a quarter to several times a second. Fidelity is how closely it matches reality, in geometry and in data.
The Consortium’s current definition page also separates digital twin systems, which use real-time and historical data from an asset that exists, from digital twin prototypes, which are modelled before the real-world counterpart is connected. That distinction is useful in property, where most twins begin life before the building does.
How is a digital twin different from BIM or a 3D model?
A 3D model shows what something looks like. A building information model (BIM) adds structured information about every element, and ISO 19650 sets out how that information is managed across the life of a built asset, from planning and design to operation. A digital twin goes one step further and connects the model to what is happening now: occupancy, energy, equipment status, construction progress.
In practice they stack. BIM is usually the best source for a twin’s geometry and asset data. The open format for exchanging it is IFC, published as ISO 16739-1:2024, whose latest edition extends beyond buildings to infrastructure such as roads, bridges, railways and ports.
What are the levels of a digital twin?
There is no single official scale, so vendors use the word for very different products. The five-level ladder below is the working model we use to scope projects. It is not a formal standard. Each level includes the ones before it.
- Level 1, visual twin: an accurate, explorable 3D model of the development. Used for sales, marketing, stakeholder approval and master planning. Updated when the design changes.
- Level 2, informational twin: every unit, floor and asset in the model carries data, such as availability, price band, specifications or maintenance records, drawn from BIM and business systems.
- Level 3, connected twin: live data flows in from sensors, the building management system and other platforms, so the model shows current conditions.
- Level 4, analytical twin: the model is used to simulate and predict, for example energy demand, crowd flow, or the effect of a design change.
- Level 5, operational twin: the link runs both ways, and operators or automated rules act on the real asset from the twin.
Most real estate projects need level 1 or 2 during sales and construction, and level 3 once the asset is operating. Levels 4 and 5 are justified at the scale of a district, a campus or a city, where small percentage gains in energy or maintenance are large in absolute terms. Buying level 4 software for a level 1 problem is the most common way these projects disappoint.
How do developers and giga-projects use digital twins?
Uses follow the life of the asset, and the audience changes at each stage.
- Planning and approvals: test massing, views, shading and phasing with investors and authorities before anything is built.
- Sales and leasing: let a buyer walk an unbuilt unit on the web, on a touch table in the sales centre or in a VR headset, with live availability.
- Construction: compare progress on site with the programme and the model, and show it to owners who are not on site.
- Operations: monitor energy, water, air quality, lifts and security from one model instead of separate dashboards.
- Community and visitors: give residents and tourists a map that knows what is open, busy or reserved.
Public policy in the region points the same way. The Dubai Metaverse Strategy, launched in July 2022, names digital twins among its pillars alongside extended, augmented, virtual and mixed reality.
What does it take to build a digital twin?
A twin is a data project with a 3D front end. The build has six parts, and the order matters.
- A purpose and an owner: one named decision the twin will improve, and one team accountable for keeping it current.
- Source geometry: BIM and CAD files for what is designed, and drone photogrammetry or laser scans for what exists. Expect clean-up, because design models are not built for real-time display.
- A real-time engine: Unity or Unreal Engine for high visual quality on headsets, screens and kiosks, or web delivery for reach. For large sites, the OGC 3D Tiles community standard is designed for streaming and rendering massive 3D geospatial content.
- Data integration: connections to sales and property systems at level 2, and to sensors, building management and IoT platforms at level 3 and above.
- A data platform: storage, history, access control and an API, so the twin is not the only place the data lives.
- An update process: who changes the model when the design or the building changes, and how often. This is the specified frequency in the definition.
What should a twin be governed by?
The Gemini Principles, published in 2018 by the Centre for Digital Built Britain, remain a good checklist. They group nine principles under three headings: purpose, meaning the twin must serve a clear public or business good; trust, covering security, openness and quality; and function, covering how twins federate, are curated and evolve.
For a developer, that translates into three questions to settle before procurement. What decision does this twin improve? Who is allowed to see which data, including data about residents? And will the model and its data stay usable if you change supplier, which is the argument for open formats such as IFC and 3D Tiles.
What does Eurisko X bring to a digital twin project?
We have no published project that connects a twin to live sensor data, and none for a real estate developer or a giga-project. What we have published are explorable 3D replicas, which is levels 1 and 2 on the ladder above: the Bank Audi VR branch, a fully designed branch Eurisko built for Bank Audi in 2022 that users walk through in VR; Advanced Supermarket VR, our own VR store, where a CMS sets the placement of every item and analytics record each interaction; Interior Design VR, our own walkthrough of furnished apartments where flooring, wall colours and materials change in real time; and architectural visualisation that turns 2D AutoCAD drawings into 3D optimised for a desktop application. Eurisko X builds this kind of replica in Unity, Unreal Engine and for the web.
The live-data half is covered by our IoT practice. In the Smart Forest project, Eurisko delivered low-power sensor nodes networked over LoRaWAN, a data pipeline built for intermittent links, and an operations dashboard showing a live map of every node with its readings, battery health and alerts. That is the same sensor-to-screen chain a level 3 twin depends on. Eurisko X is the games and immersive studio of Eurisko, founded in 2009, with more than 200 in-house staff and offices in Adma, Riyadh, Dubai and Larnaca.
Sources
- Digital Twin Consortium, definition of a digital twin (December 2020, definition page read November 2026)
- ISO 19650-1:2018, information management using building information modelling, concepts and principles
- ISO 16739-1:2024, Industry Foundation Classes (IFC) data schema
- Open Geospatial Consortium, 3D Tiles community standard, version 1.1 (read November 2026)
- Centre for Digital Built Britain, The Gemini Principles (2018)
- Government of Dubai Media Office, launch of the Dubai Metaverse Strategy (July 2022)