3D Scanning Services — LiDAR, Photogrammetry & Drone-Based Reality Capture
We mobilise LiDAR, photogrammetry and drone crews globally — across every region we serve.
What this page covers
Our 3D Scanning Modalities
What We Use 3D Scanning For
Need a 3D scan, point cloud or as-built model?
Key points covered
Survey-grade terrestrial and mobile LiDAR capture for plants, tanks, pressure vessels, pipe racks and structures. Millimetre-accurate point clouds for as-built models, clash detection, deformation monitoring and dimensional control — registered to plant coordinates and ready for CAD, BIM and digital-twin workflows.
High-resolution photogrammetric reconstruction for surface mapping, corrosion and coating documentation, weld profiling and reverse engineering. Colourised, texture-rich meshes and orthomosaics that pair with NDT data for traceable, geo-referenced inspection records.
Drone-mounted LiDAR and photogrammetry for confined, elevated and hard-to-access assets — flare stacks, storage-tank roofs, bridges, jetties, cooling towers and offshore structures. Reduces rope-access and scaffolding cost while improving inspector safety and coverage.
Tank & pressure-vessel deformation / settlement surveys (API 653, API 510)
A 3D scan produces a measured record of what is physically on site, not what the drawings say. It answers geometry questions: brownfield tie-in dimensions, turnaround access planning, tank settlement and out-of-roundness, as-built records where none exist. It does not measure wall thickness or find cracks — capture and NDT answer different questions. Atlantis NDT mobilises LiDAR, photogrammetry and drone crews globally.
A walkthrough model, a clash check against a proposed tie-in, an as-built drawing set and dimensional verification against design tolerance are four different jobs with four different accuracy requirements — scope the job, not the point density. Scanning earns its keep where geometry decisions are expensive: a connection area scanned before fabrication converts site rework into a shop dimension, and a pre-shutdown scan lets scaffolding, access and lift plans build against measured reality, which is where most turnaround schedule slip originates. Shell distortion, settlement and out-of-roundness are measurable from a scan faster than by manual survey and comparable campaign to campaign. Before commissioning, decide what condition data will attach to the geometry and at what resolution — the constraint on a digital twin programme is reconciling the corrosion monitoring location register, not capturing geometry, and whole-plant high-density capture before that decision is the most common overspend. If existing as-builts are trustworthy, re-scanning may change no decision.
Source: USIBD Level of Accuracy (LOA) Specification, Guide C120, for capture tolerance; BIMForum Level of Development (LOD) Specification for modelling detail; API 653 Annex B and API 510 where the deliverable is settlement or deformation evaluation.
Technically reviewed by Anoop Rayavarapu — ASNT NDT Level III (UT, RT, MT, PT, VT, ET) · API 653 · ISO 9001:2015 Lead Auditor
Four scanning jobs, four accuracy requirements
Use case
Deliverable
Accuracy driver
Typical method
Brownfield tie-in
Connection-area point cloud and dimensions
Fabrication tolerance of the new spool
Terrestrial LiDAR
Turnaround planning
Walkthrough model, access and lift plans
Clash margins, not millimetres
Terrestrial and mobile LiDAR
Tank and vessel geometry
Settlement, verticality, out-of-roundness
API 653 / API 510 evaluation
Terrestrial LiDAR
As-built record where none exists
Revit, IFC or DWG model set
Agreed level of development, not maximised
LiDAR plus photogrammetry
Surface condition documentation
Colourised mesh, orthomosaic
Image resolution, not geometry
Photogrammetry
Digital twin geometry
DT-ready model bound to the CML register
CML location resolution
LiDAR, scoped per unit
A scan records surfaces only — no wall thickness, no cracks, nothing internal. Pair it with NDT; the scan gives geometry, the examination gives condition.
Does a 3D scan detect cracks or wall loss?
No. A scan records surfaces — it cannot measure wall thickness, find cracks or see inside anything. Reality capture and NDT are complementary: geometry from the scan, condition from the examination.
When is re-scanning a waste of money?
When existing BIM, CAD or isometrics are trustworthy. If the as-builts already support the decision at hand, a new scan adds nothing that changes it — verify the drawings before commissioning capture.
What is the most common way to overspend on reality capture?
Scanning an entire plant at high density before deciding what condition data attaches to the geometry. The digital twin constraint is reconciling the corrosion monitoring location register, not capturing geometry — scope the scan to units where condition data justifies it.
How should a scanning contract specify accuracy and detail?
As two separate numbers: the USIBD Level of Accuracy specification for capture tolerance, and the BIMForum Level of Development specification for how much geometry gets modelled.
Where does scanning save the most on a brownfield site?
Tie-ins. Scanning the connection area before fabrication catches pipework that moved during unrecorded modifications, converting a site rework into a shop dimension.
Why scan before a turnaround?
So scaffolding, access and lift plans are built against measured reality rather than drawings — mismatches there are where most turnaround schedule slip originates.
What reality capture delivers, and what it does not
3D laser scanning, photogrammetry and drone survey produce an accurate, measurable record of an asset as it exists today. That record is the input to as-built engineering, dimensional control, clash detection, turnaround planning and digital twin construction — but it is geometry, and geometry alone answers none of the questions an integrity team is asking.
Capture methods and when each applies
Terrestrial laser scanning — survey-grade accuracy for process plant, structures and confined spaces where dimensional precision drives the deliverable.
Photogrammetry — high-resolution texture and colour, useful where surface condition and coating state matter as much as dimension.
Drone and UAV capture — elevated structures, tank roofs, flare stacks, tailings facilities and anything where access would otherwise mean scaffold or rope access.
Existing BIM, CAD and isometrics — frequently the cheapest starting point. If as-builts are trustworthy, re-scanning may add nothing that changes a decision.
Deliverables
Registered point clouds in LAS, E57, RCP or RCS with stated registration accuracy.
As-built models in Revit, IFC, AutoCAD or MicroStation, at a level of detail agreed against the use case rather than maximised by default.
Deformation and dimensional comparison against design or against a previous scan.
Geometry prepared for ingestion into a digital twin, where inspection data is then bound to locations on the model.
The decision worth making before you commission a scan
Decide first what condition data will be attached to the geometry, and at what resolution it needs to be located. Scanning an entire plant at high density before that decision is the most common way to overspend on reality capture: the constraint on a digital twin programme is almost always reconciling the corrosion monitoring location register, not capturing geometry. Scope the scan to the units where the condition data justifies it.
What 3D scanning actually delivers on an industrial site
Laser scanning and photogrammetry produce a measured record of what is physically there — as opposed to what the drawings say is there, which on a plant more than a few years old are two different things. The deliverable is a point cloud, and its value depends entirely on what you intend to do with it: a walkthrough model, a clash check against a proposed tie-in, an as-built drawing set, or dimensional verification against design tolerance are four different jobs with four different accuracy requirements.
Where it earns its keep
Brownfield tie-ins. The classic failure is a fabricated spool that does not fit because the existing pipework moved during a modification nobody recorded. Scanning the connection area before fabrication converts a site rework into a shop dimension.
Turnaround planning. Scanning before a shutdown lets scaffolding, access and lift plans be built against measured reality, which is where most turnaround schedule slip originates.
Tank and vessel geometry. Shell distortion, settlement profiles and out-of-roundness are measurable from a scan far faster than by manual survey, and the result is comparable campaign to campaign — which is what makes it evidence rather than an observation.
As-built records where none exist. Older facilities frequently have no reliable drawing set. A scan is the cheapest route to one.
What it does not do
A scan records surfaces. It does not tell you wall thickness, it does not find cracks, and it cannot see inside anything. Reality capture and NDT answer different questions and are complementary rather than alternative — the scan tells you the geometry, the examination tells you the condition. Anyone selling scanning as a substitute for inspection is selling past the physics.
Accuracy is also a specification, not a property: register the survey properly and control the target network, or the point cloud is precise and wrong. Ask any provider what registration error they achieved, not just what the scanner is rated for.