A Practical Scan-to-BIM Procurement Guide

A scan-to-BIM procurement guide is most useful before a supplier is appointed, not after a point cloud has been captured and expectations begin to diverge. The decisions made at briefing stage determine whether the resulting model supports design, planning, coordination and listed-building consent work, or becomes another file that needs checking, repairing and redrawing.

For architects and consultants, procurement is not simply about comparing a day rate for scanning or a price per square metre for modelling. It is about commissioning dependable existing-condition information at the right level of detail, with clear responsibilities for accuracy, interpretation and delivery. A well-written scope protects programme, design confidence and professional time.

Start with the decisions the model must support

The right scan-to-BIM brief begins with use, rather than software. Ask what the project team needs to know from the model in the next stage of work. A feasibility study for a commercial fit-out has different requirements from a complex refurbishment, a heritage record, or a coordination model for new services within an irregular roof space.

This distinction affects how the building is captured, what is modelled and where verification effort is focused. If the model will support proposed layouts, the team may need reliable wall positions, floor levels, openings and primary structure. If it will be used to design a new staircase, façade intervention or MEP installation, local geometry, soffit heights, service routes and structural constraints may require closer definition.

Avoid a broad instruction such as “produce a Revit model of the building”. It leaves too much open to interpretation. Instead, identify the intended uses, the design stage, the disciplines relying on the information and the areas where accuracy is especially critical. A supplier can then recommend a proportionate capture and modelling approach rather than pricing an assumption.

Define the scan-to-BIM procurement scope clearly

A useful scope separates site capture from documentation and documentation from model interpretation. Laser scanning records measured geometry. The BIM model is a structured representation of that geometry, built to agreed conventions and tolerances. They are connected, but they are not interchangeable.

State which spaces, elevations, roofs, basements, external areas and inaccessible zones are included. Clarify whether the brief covers every room, only surveyed areas, or a defined package of works. On older or occupied buildings, access assumptions deserve particular attention. Locked rooms, dense storage, live plant areas, restricted roof access and fragile historic fabric can all affect completeness.

The same applies to building elements. Specify whether the model should include primary and secondary structure, internal partitions, doors, windows, stairs, ceilings, roof forms, fixed joinery, visible services or external features. For listed properties and architecturally sensitive spaces, apparent irregularities should not be automatically simplified. Leaning walls, uneven floors, distorted openings and non-orthogonal junctions may be central to the design problem.

A good brief also identifies exclusions. Loose furniture, small decorative features, concealed structure and unverified construction build-ups are commonly outside scope unless there is a clear reason to include them. Explicit exclusions are preferable to assumptions that surface during delivery.

Specify outputs, not just a model format

A Revit file alone may not be enough. Your team may also need registered point clouds, 2D floor plans, elevations, sections, roof plans, reflected ceiling plans, PDF drawing sheets or an export for another coordination environment. Confirm the required versions, naming conventions, coordinates and units before work starts.

Consider how the model will enter the project’s wider information process. Is there an existing Revit template? Are shared coordinates available? Does the design team need separate discipline models, worksets, parameters or particular classification fields? These requirements can be efficient to accommodate when known in advance, but disruptive when introduced after modelling has begun.

For many projects, the most practical arrangement is a clean existing-condition Revit model supported by the registered point cloud and a concise drawing set. This provides both a usable design base and a measured reference for checking areas that need closer attention later.

Agree accuracy and level of development without false precision

Terms such as LOD100 to LOD400 can be useful, but they do not replace a project-specific description of what will be represented. Two models described with the same LOD can differ substantially in geometric reliability, element coverage and information content.

Set an overall accuracy requirement, then identify local tolerances where they matter most. A general arrangement model may be suitable within a modest tolerance across standard rooms, while a staircase opening, retained façade, plant room or complex roof junction needs more exact definition. The required tolerance should relate to design risk and proposed intervention, not an arbitrary aspiration for maximum detail everywhere.

It is also worth distinguishing measured geometry from inferred geometry. A scan can record visible surfaces with high precision, but it cannot reveal what sits behind a lining, above a suspended ceiling or inside a wall. Modelled assumptions should be identified where relevant. This is particularly important in renovation and heritage work, where concealed conditions can change design decisions.

Ask how irregular geometry will be handled. Will non-plumb walls be modelled as they are, represented by a best-fit plane, or documented through sections and point-cloud reference? There is no single correct answer. The appropriate method depends on whether the irregularity affects the design, the fabric’s significance and the model’s intended use.

Assess the supplier’s capture and modelling method

A capable supplier should explain their workflow in plain terms: how they plan site coverage, register scan data, check completeness, model from the point cloud and review the final outputs. The answer should demonstrate control, not unnecessary technical theatre.

For complex buildings, experience matters as much as equipment. A listed townhouse with settled floors, shallow voids and layered alterations requires different judgement from a regular new-build shell. The same is true of curved façades, vaulted ceilings, historic staircases and buildings assembled through multiple extensions.

Request relevant sample outputs where possible, including examples of drawings and models rather than visualisations alone. Look for clean geometry, sensible object structure, readable annotations and evidence that the team understands how designers will use the files. A polished image does not confirm that levels align, doors are correctly located or roof geometry can be relied upon.

You should also establish whether scanning and modelling are undertaken under one coordinated responsibility. Splitting capture and documentation between separate parties can work, but it increases the need for clear handovers and accountability. A single specialist team can often resolve site queries more directly because it understands both the source data and the final deliverable.

Build quality assurance into the appointment

Quality assurance should be visible in the scope, not treated as a final-stage promise. Ask what checks will be undertaken against the point cloud, how drawing dimensions are verified, and how omissions or access limitations are recorded. For larger or higher-risk projects, agree a review point after initial capture or an early model area before full production proceeds.

Acceptance criteria should be practical. They may cover agreed areas, required outputs, file health, coordinate alignment, tolerances and the treatment of known limitations. They should also establish a clear period for the client team to review the deliverables and raise genuine scope or quality queries.

The procurement process should not encourage suppliers to conceal uncertainty. A dependable partner will flag areas where line of sight was obstructed, access was unavailable or geometry could only be interpreted from partial evidence. That information allows the project team to arrange a return visit, investigate opening-up works or manage the risk consciously.

Compare proposals on value, not the lowest headline fee

A low fee can be attractive until missing areas, simplified geometry or unsuitable file structure consume design-team hours. When comparing quotes, check whether each proposal covers the same survey extent, modelling detail, outputs, turnaround and revision allowance. Price comparisons are meaningful only when the underlying scope is equivalent.

Programme deserves similar scrutiny. Fast delivery is valuable when it is supported by realistic site access, adequate resource and a defined review process. It is less useful when the programme assumes instant information from the client or leaves no time for quality control.

Space Captures approaches this work as precision-first documentation: measured capture, structured outputs and clear communication around what the data can and cannot confirm. That is especially valuable where irregular geometry or heritage constraints make generic assumptions costly.

Before issuing the appointment, make sure the project team can answer a straightforward question: what decisions should this model allow us to make with confidence? If the scope, accuracy, outputs and checks all support that answer, the scan-to-BIM package is far more likely to become a dependable design foundation rather than another item to resolve.