Capabilities
Four core methods and the honest boundary of each. Every procedure below has a written uncertainty budget available on request under NDA where a client dataset is involved.
Terrestrial laser scanning
Phase-shift and time-of-flight instruments on a surveyed control network. The right choice depends on range, surface reflectivity and how wet the site is — not on which instrument we happen to own.
Where it works
- As-built verification against design
- Structural deformation monitoring
- Congested plant rooms and industrial interiors
- Clash detection before a shutdown
Where it struggles
- Highly reflective or transparent surfaces
- Actively wet exteriors
- Anything moving during acquisition
- Very fine surface texture below sampling density
Typical uncertainty
- 1.8mm median registration residual at 20m
- 3–5mm on wet exteriors
- Control network tie under 2mm on a braced loop
Photogrammetry
Close-range and UAV photogrammetry where colour fidelity matters more than raw range accuracy, or where an instrument physically cannot be set up.
How we scale it
Scale comes from surveyed control, never from a scale bar alone on anything longer than a few metres. Corridor captures use the braced control placement described in method note VR-2026-02 because unbraced runs drift.
What you receive
Textured mesh, orthophotos and a dense cloud, each with the camera network report. Where the capture is intended for later re-processing, the raw imagery is delivered too.
Digital twin assembly
Turning measured geometry into an asset model somebody can actually maintain from — which means provenance, not just polygons.
Level of information need
Agreed in writing before capture. Modelling everything to the highest detail is expensive and usually wrong; we scope to the decisions the model must support.
Provenance record
Capture date, instrument, operator, control coordinates and processing chain travel with the model as structured metadata for the life of the asset.
Re-capture planning
A twin is a snapshot. We set out how often the asset should be re-captured and what would invalidate the current model.
Dimensional metrology
Where the deliverable is a pass/fail against a tolerance rather than a picture. Handled with a documented uncertainty budget, because a measurement without one is an opinion.
| Task | Method | Expanded uncertainty (k=2) |
|---|---|---|
| Fabricated steelwork verification | Laser tracker or TLS with control | 0.4–1.2mm depending on span |
| Flange and mating face flatness | Close-range photogrammetry | 0.15–0.4mm |
| Building envelope as-built | TLS on braced network | 3–6mm |
| Heritage surface condition | Structured light, local frames | 0.05–0.2mm |
Sectors
Rail, port and utilities
As-built verification, clearance envelopes and deformation monitoring, usually inside short possession windows where the capture plan matters more than the instrument.
Monuments and collections
Conservation-grade capture with re-processing in mind, delivered alongside raw data so a future team is not locked to today's software.
Fabrication and assembly
First-article verification, jig alignment and reverse engineering of legacy parts with no surviving drawings.
Film, games and simulation
Environment and prop capture where the asset has to be both visually convincing and dimensionally honest — set extensions that match the real stage.
Loss and condition record
Pre-works and post-incident condition capture, timestamped and provenance-tracked so it holds up when the record is contested.
Academic collaboration
Instrument comparison and dataset publication with university partners. Open datasets are released under CC BY 4.0.