Lascer Scanning Accuracy & Point Cloud Requirements

How Accurate Data Underpins Reliable Scan-to-BIM Modelling
Accurate BIM models begin with accurate survey data. At Tower Surveys, our laser scanning methodology is built around achieving the highest possible capture accuracy, point density and registration precision.
Registration RMS is typically maintained at ≤ 3 mm, subject to site conditions and network geometry.

This page outlines the technical standards behind our point cloud production — including scanner performance, density expectations, registration methodology, accuracy tolerances, and georeferencing requirements. These factors directly determine the reliability and build quality of the as-existing BIM model.

Why Point Cloud Accuracy Matters

A BIM model is only ever as good as the point cloud it is derived from.
High-quality scan data enables:

  • Correct geometry and measurements
  • Proper positioning and alignment of building elements
  • Reduction of design risk
  • Improved clash detection
  • Better coordination
  • More accurate costing
  • More reliable asset information

If the point cloud is inaccurate, noisy or incomplete, the model will reflect those weaknesses, which leads to downstream issues for the entire project team.

Our approach provides clarity, precision and consistency through every stage of the data capture process.

Why Point Cloud Accuracy Matters

1. Laser Scanners & Expected Accuracy Levels

We deploy industry-leading Leica terrestrial laser scanners selected to suit project scale, geometry and required tolerances. 

Typical achievable site accuracy:

  • Relative accuracy: ±5–7 mm RMSE
  • Absolute accuracy: ±10–15 mm RMSE

Actual performance is dependent on scan geometry, control strategy, environmental conditions and registration methodology.

Registration RMS is typically maintained at ≤ 3 mm, subject to site conditions and network geometry.

We select the appropriate scanner depending on:

  • Required tolerance
  • Level of Detail (LOD)
  • Level of Information
  • Project type (heritage, industrial, commercial, residential)
  • Access conditions
  • Range and complexity

Combining the right hardware with our structured end-to-end Scan-to-BIM workflow ensures consistent and measurable results.

2. Point Cloud Density Requirements

The density of points captured determines the level of detail that can be confidently modelled.

Typical point density targets:

Standard Architectural Modelling

  • 6 mm @ 10 m
    Suitable for LOD3 modelling of walls, floors, openings, roofs and general building fabric.

High-Detail or Heritage Recording

  • 3 mm @ 5–10 m
    Used for intricate features such as:
  • Stonework
  • Mouldings
  • Decorative façades
  • Historic staircases
  • Vaults and arches

Industrial/Mechanical Environments

  • Higher density around pipework and plant
  • Supplementary scans for congested or complex areas
  • Ensures accurate MEP modelling and clash detection

By controlling density, we maintain predictable modelling tolerances across all disciplines.

3. Scan Coverage & Line-of-Sight Strategy

Quality of coverage is just as important as density and accuracy.
Our team ensures:

  • Every room and space is captured from multiple viewpoints
  • Overlapping scans ensure high registration confidence
  • Obstructions (furniture, machinery, racking, plant) are scanned around where possible
  • Staircases, voids and upper levels are fully documented
  • External façades include adjacent context where needed

We always aim to eliminate “blind spots” so the modeller does not need to make assumptions.

4. Registration Methodology

Once scans are collected, individual scans must be aligned (“registered”) into a single coordinated dataset.

Our registration methods include:

Target-Based Registration

Using reflective or spherical targets placed throughout the site.
Best for:

  • Complex interiors
  • Multi-level spaces
  • Heritage projects
  • Environments with repetitive geometry

Cloud-to-Cloud Registration

Aligning scans based on overlapping geometry, without physical targets.
Best for:

  • Open spaces
  • Industrial units
  • Large externals

Hybrid Registration

Combining targets and cloud matching for maximum accuracy.

Registration Verification

We verify registration by checking:

  • Residuals between scans
  • Mean and RMS errors
  • Overlap consistency
  • Visual alignment checks

Our typical target RMS is below 3 mm, depending on site conditions and scanner type.

Registration Methodology

5. Noise Reduction & Point Cloud Cleaning

Raw point clouds include stray points, reflections, unwanted artefacts and environmental noise.
Before modelling, we apply:

  • Noise filtering
  • Removal of stray points
  • Removal of duplicated geometry
  • Removal of moving objects (people, vehicles, machinery)
  • Sharpening of edges through advanced processing
  • Smoothing of surfaces (where appropriate)

This ensures the modeller works from a refined, accurate dataset.

6. Coordinate Systems & Georeferencing

We provide point clouds aligned to:

  • OS National Grid
  • ODN (Ordnance Datum Newlyn)
  • Local control grids
  • Client-supplied coordinates
  • Shared Revit coordinates for multi-disciplinary collaboration

Consistent coordinate alignment ensures seamless integration with:

  • Architectural design teams
  • Structural engineers
  • MEP designers
  • Contractors
  • BIM coordination environments
  • GIS systems
  • Topographical surveys

Georeferencing is performed either during registration or as part of a post-processing step using supplied survey control.

7. Accuracy Tolerances & Expected Outputs

Our typical tolerances:

Point Cloud Accuracy

  • Single-point measurement capability is typically in the 1–6 mm range (depending on scanner and range). Overall project accuracy is governed by scan geometry, control and registration.
  • Registration RMS typically ≤ 3 mm

Model Tolerances (linked to LOD)

  • LOD2: Approx. 50 mm
  • LOD3: 15–30 mm
  • LOD4: 0–15 mm

These tolerances are always confirmed during the briefing stage and documented in the specification. Our QA checks are also based on point cloud accuracy.

8. Limitations & Obscured Geometry

Laser scanners must see a surface to record it.

Examples of limitations:

  • Behind furniture
  • Above ceilings
  • Behind cladding or wall linings
  • Beneath raised floors
  • Concealed voids
  • Dense vegetation
  • Machinery or storage blocking access

If areas are obscured, we will:

  • Report them clearly
  • Annotate drawings/models
  • Discuss alternatives (selective strip-out, night access, elevated scanning, drone data, pole-mounted scanning, etc.)

Transparency prevents assumptions and errors during modelling.

9. Delivering Point Cloud Data

Point clouds are supplied in one or more of the following formats:

  • RCS / RCP (Autodesk ReCap / Revit preferred)
  • E57
  • PTS / PTX
  • LAS / LAZ (for externals)
  • PLY (for meshes or specialist workflows)

We can also provide:

  • TruView / 360° imagery
  • Orthophotos
  • Slices and sections
  • Mesh conversions (OBJ, FBX, STL)

These formats support a wide range of BIM, CAD and coordination environments.

Delivering Point Cloud Data

Summary

Laser scanning accuracy, point density, registration quality and georeferencing form the technical backbone of any Scan-to-BIM project.
Our structured approach ensures:

  • Registration RMS is typically maintained at ≤ 3 mm, subject to site conditions and network geometry.
  • High-fidelity as-existing data
  • Reliable BIM geometry
  • Predictable tolerances
  • Fully verifiable outputs and survey deliverables

With the right data, everything downstream becomes faster, safer and more accurate.

Return to the full Scan-to-BIM Survey Specification

Need a Quote or Have a Project to Discuss?

We provide rapid quotations and can advise on scope, deliverables, LOD levels and expected workflows.
Contact us today to discuss your Scan-to-BIM requirements.
This site is protected by reCAPTCHA and the Google Privacy Policy and Terms of Service apply.
Tower Surveys 3D Modelling logo
Tower Surveys 3D Modelling Ltd are part of Tower Surveys Associates Group
Company number 14623925,
Registered office address:
42 Deerness Heights,
Stanley, Crook,
England, DL15 9TJ

Copyright Tower Surveys 3D Modelling 2025 | Privacy Policy