Quick Answer
Scan to BIM (or Scan to HBIM for heritage) is the process of using point cloud data captured by laser scanning or photogrammetry as the geometric reference for building a parametric BIM model of a historic structure. The point cloud is imported into BIM software (typically Autodesk Revit) and used as a snap reference for modelling walls, columns, roofs and architectural elements, each carrying attribute data about material, condition, period and significance.
The standard scan to BIM workflow — capturing a building's as-built geometry with a laser scanner, then building a parametric model from that data — was developed for new or recent buildings where the point cloud reveals a structure built to regular tolerances. Applying it to historic buildings, where nothing is regular and the irregularity is part of the significance, requires adapting the approach at every stage.
This adaptation is worth making. The HBIM model that results from a well-executed scan to HBIM process is a management tool of lasting value: it can be queried by element, updated after each inspection, linked to conservation decision records, and used to generate accurate drawings and quantities for repair specifications. The effort invested in creating it returns value across many subsequent conservation and management activities.
What is Scan to BIM?
Scan to BIM is the process of using point cloud data captured by laser scanning (or photogrammetry) as the geometric reference for creating a Building Information Model. In the as-built context, it is used to create accurate digital representations of existing structures — both new buildings where the as-built condition may differ from the design intent, and historic buildings where no design documentation exists.
The point cloud is imported into BIM authoring software and used as a visual and geometric reference for model creation. Each BIM object — a wall, a column, a floor slab — is modelled to match the geometry of the corresponding point cloud feature. The resulting model combines the geometric accuracy of the scan data with the parametric, attribute-carrying structure of BIM.
Scan to BIM vs Scan to HBIM
Standard scan to BIM, applied to recent buildings, typically produces a model with regular elements at standard tolerances, referencing standard material specifications and construction types. The workflow is relatively straightforward because the building was designed and built to regular dimensions.
Scan to HBIM for historic structures differs in several important ways. The geometry is irregular — walls that are not plumb, columns whose diameters vary, arches that were set out by approximate geometric construction rather than by compass and rule. The construction materials are non-standard — rubble masonry, lime mortar, timber of irregular section. The elements require attribute data that standard BIM objects do not carry — period attribution, condition grading, significance level, repair history. And the purpose of the model is conservation management rather than construction coordination, which changes what information is important to capture.
Scan to BIM vs Scan to HBIM
| Aspect | Scan to BIM (new buildings) | Scan to HBIM (historic buildings) |
|---|---|---|
| Geometry | Regular, standardised | Irregular, non-standard |
| Materials | Standard spec materials | Historic, non-standard |
| Element libraries | Standard BIM families | Custom heritage families |
| Key attributes | Specification, cost, schedule | Condition, period, significance |
| Primary purpose | Construction coordination | Conservation management |
| Accuracy required | ±10–20 mm typical | ±5–10 mm typical |
The Scan to HBIM Workflow
Stage 1: Survey data capture. Point cloud data is captured by TLS or photogrammetry, processed, registered and quality-checked. This stage must achieve the accuracy and completeness the HBIM model will require — gaps in the point cloud cannot be filled in modelling.
Stage 2: Point cloud import. The registered, colour point cloud is imported into the BIM environment (Revit via Autodesk ReCap, ArchiCAD, BricsCAD). Point cloud density is balanced against file performance; thinned subsets are used for reference during modelling, with the full-density cloud available for verification.
Stage 3: Horizontal and vertical datum establishment. Key reference levels (floor levels, datum elevation) and vertical datum planes are established from the point cloud and used to orient the model coordinate system.
Stage 4: Custom family creation. For non-standard elements — traditional column types, corbelled brackets, carved capitals — custom parametric Revit families are created. This is typically the most time-consuming part of the workflow for complex historic structures.
Stage 5: Element modelling. Working from the point cloud reference, each building element is modelled to match the observed geometry. Attribute data is assigned to each element as modelling proceeds.
Stage 6: Accuracy verification. Model geometry is checked against the point cloud at representative points. Deviations are documented.
Stage 7: Attribute data completion. Condition grades, period attributions, significance levels and maintenance notes are completed and reviewed.
Stage 8: IFC export and delivery. The model is exported to IFC for interoperability, delivered with model documentation.
Handling Geometric Irregularity
The most significant modelling challenge in Scan to HBIM is representing geometric irregularity accurately while maintaining the parametric, queryable structure of a BIM model. Several approaches are used, each with trade-offs.
For walls that are not plumb: in Revit, wall objects can be modelled as sloped or tapered to match the point cloud. Alternatively, a series of planar wall panels approximating the actual wall surface at different heights can represent a surface that curves in plan or tilts in elevation.
For columns that are not cylindrical: custom family geometry can accommodate the specific profiles documented in the point cloud. For series of similar columns with individual variations, a master family with adjustable parameters allows each column instance to be individually adjusted to match the survey data.
For complex arches and vaulted ceilings: in-place mass modelling tools in Revit allow the creation of complex curved surfaces that do not conform to standard parametric family shapes. These in-place elements sacrifice some parametric flexibility but allow accurate geometric representation of complex forms.
The practical decision at each element type is: how much geometric accuracy is required for the intended use of the model? A structural assessment model needs more accurate geometry than a maintenance management model. Defining the purpose of the model before beginning — not after — guides these decisions.
Level of Development for Heritage
Level of Development (LoD), borrowed from standard BIM practice, defines how precisely each element in the model represents the actual building element. For heritage work, LoD is adapted to reflect both geometric precision and information richness.
LoD 200 for heritage: elements are represented as approximate shapes and sizes, not verified against survey data. Appropriate for preliminary models and site overviews.
LoD 300 for heritage: elements are modelled to sizes and locations verified against the survey data, within a tolerance of approximately 50 mm. Material types and construction periods are attributed. This is the standard level for heritage management models.
LoD 350 for heritage: as LoD 300 but with interface conditions between elements explicitly modelled. Condition grades and significance levels are attributed to all elements.
LoD 400 for heritage: element geometry is modelled to fabrication-level accuracy. All attribute data is complete. Appropriate for research-grade documentation of significant individual elements.
The appropriate LoD for a project should be defined in the brief, not left to the modeller's judgement. Higher LoD means more modelling time and higher cost; choosing LoD 400 where LoD 300 would serve the purpose wastes resources.
Software and Tools
Autodesk Revit + Autodesk ReCap is the standard combination for scan to HBIM work. ReCap processes and manages the raw scan data; the point cloud is imported into Revit through ReCap's point cloud format. Revit's parametric modelling tools, family editor and IFC export capability make it the most complete platform for HBIM work.
ArchiCAD with Pointcloud Import add-on is a viable alternative, particularly for projects where complex curved geometry is central and Revit's difficulty with non-regular forms is a constraint.
CloudCompare is a free, open-source point cloud processing tool widely used for scan registration, comparison and cleaning before import into BIM environments. Its ability to compute cloud-to-cloud distance comparisons — comparing model geometry against point cloud — makes it valuable for model accuracy verification.
Leica Cyclone and Faro Scene are the major manufacturer-supplied point cloud processing applications, used for scan registration and basic cloud processing before handover to the BIM environment.
Key Takeaways
- 1Scan to HBIM uses point cloud data from laser scanning or photogrammetry as the geometric reference for creating a parametric heritage BIM model.
- 2Unlike standard scan to BIM, HBIM must accommodate geometric irregularity, non-standard construction materials, and heritage-specific attribute data (condition, period, significance).
- 3Custom parametric families are typically required for traditional architectural elements that standard BIM libraries cannot represent accurately.
- 4Level of Development (LoD) should be defined in the project brief: LoD 300 is appropriate for most heritage management models; LoD 400 for research-grade documentation.
- 5Autodesk Revit + ReCap is the dominant platform; ArchiCAD handles complex curves better; CloudCompare is valuable for open-source point cloud processing and accuracy verification.
Frequently Asked Questions
How long does scan to HBIM take for a typical heritage building?
A modest single-storey historic building at LoD 300 might take 80–150 hours of modelling time after the point cloud is delivered. A complex multi-storey structure with elaborate architectural ornament at the same LoD might require 400–800 hours. Creating custom parametric families for non-standard traditional elements is consistently the most time-consuming component.
What happens when the point cloud has gaps?
Point cloud gaps — areas not covered by the scan — translate into model uncertainty. Standard practice is to note the gap in the model documentation, to model the missing area based on the best available information (symmetry with documented areas, historic photographs, similar documented elements), and to clearly flag the modelled element as based on inference rather than direct survey data. A return to site to fill significant gaps is always preferable to modelled inference for important elements.
Further Reading
Jabendra Raja
Technical-Commercial Partner, Evergreen Origins
Jabendra Raja leads the Technical-Commercial practice at Evergreen Origins, with experience in heritage documentation and spatial technology across South India.