Scan-to-BIM Modeling Services: What Is Modeled and What Is Not

A scan-to-BIM model does not automatically contain every object the scanner saw. It contains the categories you scoped, modeled to the level of detail you agreed, and nothing else. The walls, floors, doors, columns, ceiling grids, and visible MEP mains usually make it into the Revit model. Furniture, small conduit, decorative trim, manufacturer-specific equipment, and anything hidden behind a finish usually do not, unless you name them. The phrase “scan-to-BIM” is useful, but only once the scope is written down: which categories, at what LOD, with which exclusions, in which file formats, and how the model gets reviewed.

The reason this matters is simple. The scan gives you measured evidence. The BIM model is a set of decisions about what to represent from that evidence. If the proposal does not name what is modeled and what is excluded, one of two things happens. The quote looks cheaper than the real project because half the work is unstated, or the finished model is missing the exact thing your design team assumed was included. Both are avoidable with a few sentences of scope.

A scan is measured data; a model is interpreted geometry

A registered point cloud (delivered as E57, or as RCP/RCS for Autodesk workflows) is a dense set of measured points. It shows a wall face, a column edge, a pipe run, a doorway, a roof plane. It does not contain “a wall” as an object. A modeler creates the Revit wall by deciding where the best-fit plane sits, which construction phase it belongs to, and how much surface variation to keep.

That decision is where scope lives. A scanned wall is often wavy, out of plumb, or patched across several renovations. The model can represent a clean best-fit plane, or it can chase every bow and bulge in the as-built surface. The first is fine for a renovation background; the second is what you want for a facade restoration or a tight tolerance fit-out, and it costs more. The right answer is whichever one matches how you will use the model, which is why scope should start there.

Start with the project use, not the category list

Before listing categories, answer one question: what does this model need to support? A concept-stage test fit does not need the content a detailed renovation drawing set needs. An architecture-only background does not need overhead systems, while MEP routing coordination is mostly about the overhead systems. Demolition planning, structural review, clash detection, and owner record documentation each pull the scope in a different direction.

Once the use is clear, the category scope writes itself. For architecture, that typically means exterior walls, interior partitions, doors and windows, floor slabs, stairs, ramps, shafts, major openings, ceiling grids, and roof surfaces visible from the scan. Even here, decide the edge cases in advance: are baseboards and soffits modeled, are wall finishes separated from core, are sloped floors modeled true or simplified, are historic profiles modeled as real geometry or generic placeholders. Each of those is a small request that quietly changes modeling hours.

Structural and MEP scope follow visibility

Structural scope is governed by what the scanner can actually see. Visible columns, exposed beams, trusses, braces, and slab edges can be modeled from scan data. Framing concealed behind finishes cannot be confirmed by laser scan unless it is exposed, opened, or backed by other records. A clean structural scope says: model visible columns and beams, model exposed bracing and slab edges where visible, exclude concealed framing, reinforcement, and connection plates unless requested, and use generic structural families unless real ones are provided. If the project needs structural verification, treat the model as measured existing-conditions support, not as a substitute for engineering investigation.

MEP is where scope gets expensive fastest. Visible duct trunks, large pipe mains, air handling units, pumps, panels, and major equipment are usually worth modeling. Small conduit, flex lines, insulation, valves, hangers, and fittings multiply the hours, and concealed routing cannot be modeled at all without access. A defensible MEP scope names which systems are in, the minimum duct or pipe size that gets modeled, whether fittings and insulation are included, whether equipment is generic or detailed, whether system names and connectors are required, whether the space above the ceiling was actually scanned, and what the modeler should do where a line disappears into a wall. For most renovations, selected visible MEP is enough. Coordination-heavy work needs more, and the proposal should say which one you are buying.

Set LOD by category, not for the whole building

One of the most common scoping mistakes is asking for “a LOD 300 building model” and assigning a single label to everything. That is how you end up paying LOD 300 prices for furniture you did not want and getting LOD 200 ceilings you assumed were detailed. LOD (level of development, often paired with USIBD level-of-accuracy language for as-built work) is far more useful applied per category.

A category-by-category table like this is the single most effective thing you can put in a proposal. It removes the argument about what “LOD 300” meant before the argument happens. For the underlying definitions, the scan-to-BIM LOD guide and our breakdown of LOD 200 vs LOD 300 for scan-to-BIM are worth reading before you sign anything.

What is excluded by default

Unless the scope says otherwise, assume these are not in the model: chairs, desks, boxes, loose equipment, signage, cords, plants, movable partitions, small conduit, hidden framing, wall cavities, material layers, manufacturer data, and fabrication-level detail. Any of them can be modeled when it genuinely matters, but each has to be named. “Model the loose lab equipment in Room 214” is a fine line item; assuming it is included because the scanner captured it is how change orders are born.

Separate corrections from added scope

A good workflow includes model review, and review needs limits too. Define when the first review happens, who gives comments, how many rounds are included, whether the registered point cloud ships with the model so the client can check against it, and what format comments should take. The point of review is to catch missing scope, questionable interpretation, and usability problems, not to become an open-ended expansion of the model.

This is also where scope creep starts, always with a reasonable-sounding request: “can you add the ceilings,” “can you include the mechanical room,” “can you model the pipes here,” “can you add sheets too.” Each one may be valid, and each one changes the deliverable. The clean way to manage it is one distinction: if a modeled wall does not match the scan, that is a correction and it is on the modeler. If a category was excluded and now needs modeling, that is added scope and it gets priced. A simple comment log keeps both sides honest, and it is the difference between a model that lands on budget and one that quietly drifts.

A scope paragraph you can actually use

Here is what a workable modeling scope looks like in one paragraph. “Provide scan-to-BIM modeling from registered laser scan data. Deliver a Revit model for renovation design. Model architectural walls, floors, doors, windows, stairs, ramps, columns, major openings, roof surfaces visible from scan data, and ceiling grids in marked areas. Model visible structural columns and beams. Model visible MEP mains larger than 2 inches. Exclude concealed systems, furniture, loose equipment, small conduit, manufacturer-specific families, and asset data. Two review rounds included. Deliver RVT plus the registered E57 and RCP point cloud.” That gives a modeling team something they can price and produce without guessing.

If you want help turning a building and a use case into a scope like that, request a quote and we will scope the categories, LOD, and exclusions with you before any modeling starts. WeAre Capture handles the field capture on a Trimble X7, and the BIM modeling is produced and reviewed against that data, so the scope you sign is the model you get. See our scan-to-BIM services for the full deliverable list, or point cloud to Revit model: what is included for a closer look at the file-by-file output.

FAQ

What is scan-to-BIM?
Scan-to-BIM is the process of turning a registered 3D laser scan of an existing building into a Building Information Model, usually a Revit (.rvt) file. The scan supplies measured geometry; a modeler interprets it into walls, floors, structure, and selected systems at an agreed level of detail. It is the difference between having measured data and having an intelligent model you can design against.

How do I import a point cloud into Revit?
Convert the registered scan to Autodesk’s indexed format (RCP, which references RCS scan files) using ReCap, then in Revit use Insert > Point Cloud and link the RCP. Set the positioning to Auto - Origin to Origin or shared coordinates so the cloud lands where your project base point expects it. The linked cloud then becomes the tracing reference the modeler builds the geometry against. See how to import a point cloud into Revit for the step-by-step.

How much does scan-to-BIM cost?
There is no single dollar-per-square-foot rate and no neutral industry benchmark; price tracks element density and LOD, not floor area. As a vendor-derived US range, turnkey scan-to-BIM commonly runs roughly $0.50 to $3 per square foot for basic LOD 200 to 300 work and $3 to $10+ per square foot for high-detail MEP. Field scanning on its own is typically billed by the day. Our scan-to-BIM cost factors guide explains what moves the number.

What is the difference between scan-to-CAD and scan-to-BIM?
Scan-to-CAD produces 2D drawings or 3D linework (DWG) from the scan: plans, sections, elevations. Scan-to-BIM produces an object-based 3D model (RVT, or IFC for exchange) where walls and equipment carry information, not just lines. Choose CAD when you need drawings, BIM when you need a model to design and coordinate against. See scan-to-CAD vs scan-to-BIM.

Does the scan-to-BIM deliverable include the point cloud?
It should, if you ask for it. Receiving the registered point cloud (E57 and/or RCP/RCS) alongside the model lets your team verify the geometry against the source data during review. Whether the cloud is included is a scope decision, so name it in the proposal rather than assuming.

---

Last reviewed: May 2026.