Digital twin technology and reality capture: how laser scanning builds one
Key takeaways
- A digital twin is a virtual model connected to its physical counterpart by data, not just a 3D drawing.
- Reality capture with terrestrial and SLAM laser scanners supplies the accurate as-is geometry most twins of buildings, plants and infrastructure start from.
- Survey control from GNSS or a total station ties scans to real coordinates so the twin lines up with design and GIS data.
- The value of a twin depends on how often its data is refreshed, so plan rescans and sensor updates from day one.
- Start with one asset and a clear use case before scaling to a full digital twin platform.
What is digital twin technology?
Digital twin technology is the practice of building a virtual replica of a physical asset, process or system and keeping it connected to data from the real thing. The asset might be a pump, a production line, a building, a bridge or an entire factory. What makes it a twin rather than a static model is the link: when the physical asset changes, the digital model is updated.
So what does digital twin mean in practice? Most definitions share three parts: the physical asset, the digital model (geometry plus information), and the data connection between them. A physical asset digital twin of a plant room, for example, combines measured 3D geometry with equipment tags, maintenance records and live readings.
Levels of digital twin maturity
- Descriptive twin – an accurate 3D digital twin of the as-is condition, used for visualisation, planning and coordination.
- Informative twin – geometry linked to asset data such as serial numbers, inspection history and documents.
- Predictive or real time digital twin – a model fed by sensors and analytics that can forecast behaviour and support decisions.
Many useful projects stop at the first or second level. The digital twin concept does not require live sensors to deliver value; an accurate, current model alone can save weeks of site visits.
How does a digital twin work?
A digital twin works by layering data on top of an accurate spatial model and refreshing that data on a schedule or continuously. The digital twin architecture is usually described in four layers.
| Layer | What it contains | Typical sources |
|---|---|---|
| Physical layer | The real asset and its environment | Buildings, machines, networks |
| Data capture layer | Geometry, condition and operating data | Laser scanners, survey instruments, digital twin sensors, IoT devices |
| Model layer | 3D geometry, asset hierarchy, relationships | Point clouds, meshes, BIM models (IFC), GIS layers |
| Application layer | Dashboards, digital twin analytics, simulation | Monitoring, maintenance planning, what-if studies |
In a digital twin IoT setup, sensors stream values such as temperature, vibration or flow into the model layer, where each reading is attached to the correct object. Digital twin data analytics then compares current behaviour with expected behaviour and flags anomalies.
Geometry changes more slowly than sensor values, so it is refreshed by periodic reality capture instead of streaming. A plant may be rescanned after each shutdown, while a construction site may be captured weekly.
Digital twin vs simulation: what is the difference?
A simulation models how something could behave under assumed conditions; a digital twin represents a specific real asset and is updated with its actual data. The two are complementary, and digital twin simulation is a common use of a mature twin.
| Aspect | Simulation | Digital twin |
|---|---|---|
| Subject | A design or generic system | One specific physical asset |
| Data | Assumed inputs | Measured geometry and operating data |
| Lifespan | Usually one study | Whole asset life cycle |
| Updates | Rerun manually | Refreshed from the real asset |
Digital twin modeling and simulation work best together: the twin supplies realistic starting conditions, and the simulation tests changes before they are made on the real asset. This is the logic behind digital twin for production optimization, where a line layout or process change is tried virtually first.
What is reality capture and why does a 3D digital twin need it?
Reality capture is the measurement of existing physical conditions in 3D, usually as a point cloud, mesh or textured model. Without it, a twin of an existing asset relies on old drawings, which rarely match what was actually built.
Several reality capture technologies are used, often in combination:
| Technology | Best for | Typical accuracy |
|---|---|---|
| Terrestrial laser scanning (TLS) | Plants, buildings, structural detail | Millimetre-level per scan; a few mm after registration |
| Handheld SLAM scanning | Corridors, stairs, fast 3D mapping of large interiors | Around 1–3 cm typical |
| Photogrammetry (camera or drone) | Facades, roofs, sites, textures | Centimetre-level with good control |
| GNSS RTK and total station | Control points and georeferencing | Millimetre to centimetre |
Terrestrial laser scanning delivers the density and accuracy needed for pipes, steel and equipment. SLAM 3D mapping, explained in our guide to SLAM scanning and mobile mapping, trades some accuracy for speed. A 3D mapping camera adds colour and texture. Survey control ties everything to one coordinate system so the twin aligns with design and GIS data.
The reality capture process: creating a digital twin step by step
Creating a digital twin from reality capture follows a predictable sequence. Skipping the planning and control steps is the most common reason a model has to be redone.
- Define the use case. Decide what decisions the twin will support and the level of detail (LOD) and accuracy needed.
- Establish control. Set targets or control points and measure them with GNSS RTK or a total station.
- Capture. Scan with TLS for critical areas and SLAM for circulation spaces; add photos where texture matters.
- Register and clean. Combine scans into one point cloud, remove noise, and check registration residuals against control.
- Model. Convert the point cloud into intelligent objects, often through a scan to BIM workflow, and export in open formats such as E57, LAS or IFC.
- Link data. Attach asset IDs, documents and sensor feeds to the modelled objects.
- Maintain. Schedule rescans and data updates so the twin stays current.
Digital twin creation is an ongoing process, not a single deliverable. The capture plan should state how and when each area will be refreshed.
Digital twin examples and applications
Digital twin applications span manufacturing, process plants, construction and infrastructure. The common thread is an expensive physical asset where errors or downtime cost more than measuring it properly.
- Digital twin manufacturing – a factory digital twin records machine positions, conveyors and services, so new equipment can be placed and checked for clashes before installation.
- Process digital twin – in the digital twin process industry (oil and gas, chemicals, power), scanned piping and equipment support retrofit design, maintenance planning and operator training.
- Reality capture construction – jobsite reality capture compares as-built work with the design model to catch deviations early and verify progress.
- Infrastructure – bridges, tunnels and rail assets are scanned to track deformation and plan inspections.
- Facilities – building owners use scanned models for space planning and maintenance.
These digital twin examples show that geometry is the foundation. Digital twin technology in manufacturing, for instance, depends on knowing exactly where every machine and service run sits before simulating a change.
Reality capture software and 3D mapping software: what to look for
Reality capture software turns raw scans into usable data; 3D mapping software and digital twin platforms then organise and share it. Rather than chasing the best 3D mapping software by name, evaluate tools against your workflow.
- Registration – target-based and cloud-to-cloud registration with clear accuracy reports.
- Open formats – import and export of E57, LAS/LAZ and IFC to avoid lock-in.
- Scale – handling of billions of points, or reality capture cloud processing for very large projects.
- Modelling tools – pipe, plane and steel fitting for faster scan-to-model work.
- Sharing – web viewers so non-specialists can measure and comment.
- Integration – APIs or connectors for linking asset and sensor data in a digital twin platform.
Our article on point cloud processing covers registration and deliverables in more detail.
Benefits of digital twin technology, limits and cost
The main benefits of digital twin technology are fewer site visits, fewer design clashes and better maintenance decisions. Its limits are cost, data management effort and the need to keep the model current.
Benefits
- Remote measurement and inspection from an accurate as-is model.
- Clash detection before fabrication, reducing rework.
- A single reference for operations, maintenance and engineering teams.
Limits
- A twin that is not updated quickly becomes misleading.
- Large point clouds need storage, processing power and trained staff.
- Linking IoT data requires clean asset hierarchies and naming.
Reality capture price depends on area, detail and deliverables. Reality capture services are usually priced per day or per area, while buying equipment shifts cost to capital plus training. Terrestrial scanners typically cost several times more than handheld SLAM units, so many teams use both. For hardware, our 3D laser scanner buying guide explains the trade-offs.
Sourcing reality capture hardware for digital twin projects
Mars Geosystems supplies the scanning and survey hardware that a digital twin development team or service provider needs. The terrestrial 3D laser scanner captures up to 2 million points per second with up to 266 MPx colour, while the handheld SLAM scanner (925 g host, 0.1–70 m range) covers large interiors quickly. Both can be ordered from one unit, with OEM options for distributors. To compare options for your project, request a quote with your scan areas and accuracy needs.
Frequently asked questions
What is digital twin technology in simple terms?
Digital twin technology is a virtual copy of a real object, building or process that stays connected to real-world data. The model shows the asset's actual shape and condition, and it is updated when the asset changes. Engineers use it to inspect, plan and test changes without being on site, and to predict problems before they happen.
How does digital twin technology work?
It works by combining an accurate 3D model, usually built from laser scanning or photogrammetry, with information about the asset and data from sensors or inspections. Software links each data point to the right object in the model. Analytics and simulation then use this connected model to monitor performance, compare actual against expected behaviour and test what-if scenarios.
What is the difference between a digital twin and a 3D model?
A 3D model is a static representation of geometry, often created once for design. A digital twin represents one specific real asset and is kept connected to it through updated measurements, asset records and sensor data. A 3D model can become the geometric core of a twin, but without the ongoing data link it remains just a model.
Which reality capture technology is best for a digital twin?
There is no single best option. Terrestrial laser scanning gives the highest accuracy for plants, structures and services. Handheld SLAM scanning is faster for corridors and large interiors at slightly lower accuracy. Drone photogrammetry suits roofs and sites. Most projects combine two or more, tied together with GNSS or total station control points.
How accurate does a digital twin need to be?
It depends on the use case. Facility management and space planning can work with a few centimetres of accuracy. Retrofit design of piping, steel or equipment usually needs a few millimetres so new parts fit first time. Define the required accuracy and level of detail before capture, because it drives both the equipment choice and the cost.
What are the benefits of a digital twin in manufacturing?
In manufacturing, a factory digital twin lets engineers plan new lines, place machines and route services against the real layout, avoiding clashes during installation. Linked to production data, it can highlight bottlenecks and support production optimisation. It also gives maintenance teams a shared reference for locating equipment and history.
Need equipment for this work?
Tell us the job, the country and the quantity. An engineer replies with a configuration and price within one business day. MOQ is one unit; OEM is available.
