The question behind every deformation survey
Someone notices something. A door that no longer closes cleanly, a crack that looks longer than it did, a bearing wearing on one side. The question that follows is always the same: is this moving, and does it matter?
Answering it requires two measurements separated by time. Most assets only have one, taken after the concern was raised, which establishes the current shape and nothing about the trend. The single most valuable thing an operator can do about structural movement is capture a baseline before there is a problem — because the baseline is the thing that cannot be recovered later.
Why fixed monitoring targets are not enough
The traditional approach places permanent targets at chosen locations and reads them periodically. It is precise, it is cheap to repeat, and it has one structural weakness: it reports movement only where the targets are.
That is fine if you already know where the asset will move. Often you do not. Deformation is frequently distributed rather than local, and its pattern — which side, which end, whether it is uniform or twisting — is usually more diagnostic than its magnitude at any single point. A target grid can miss a whole-structure pattern entirely while faithfully reporting the points it watches.
Full-surface capture measures everything within line of sight, so the pattern emerges from the data rather than from where somebody decided to put a target ten years ago.
What a baseline actually is
A baseline is not simply an early scan. It is a scan tied to a control network robust enough that a later survey can be registered into the same frame and compared with confidence.
This is the part that gets skipped, and it is the part that determines whether the exercise works. If the first survey is registered only to itself, a comparison five years later measures the difference between two coordinate systems as much as any real movement. Control is what makes the two datasets comparable.
It is also why baselines are worth capturing opportunistically. If an asset is being scanned for as-built modelling or calibration, extending the scope to establish a proper control network costs relatively little and creates an option you may be very glad of later.
What the comparison produces
With two registered datasets, the deviation analysis produces a surface-to-surface comparison across the whole captured structure, normally presented as a colour deviation map with measured sections through the areas that matter.
Our RRS Discovery survey in Dundee is a clean example, because the two surveys are six years apart. The 2014 scan was commissioned as a datum. The 2020 re-survey was compared against it, and the comparison returned typical section deviations along the hull of 10.4 mm, 12.7 mm, 15.4 mm, 21 mm and 35.7 mm, with an overall lengthwise extension of 34 mm.
What matters is not any single figure. It is that the movement was asymmetric and favoured the port side — a whole-hull pattern that only appeared because the whole hull was measured, twice, in three dimensions. That let the conservation team discuss a rate rather than an impression, target investigation where the movement was concentrated, and set the timing of the next re-survey knowing what magnitude of change the method resolves.
Applying it to industrial assets
The principle transfers directly. Storage tanks settle differentially and go out of round. Rotating equipment loses alignment across roller bands and support tyres. Structural steel moves under sustained load or after modification. Jetties, quay walls and support structures shift over years.
In each case the useful output is the same: a described pattern of movement with magnitudes attached, against a known reference. On rotating plant, damage and deformation assessment typically extends to fitting best-fit geometry through the structure — planar sections at regular intervals to assess roundness, cylinders fitted through roller bands to establish whether they share a common axis, centreline analysis to identify bow or step changes.
One practical caution from doing this work. Local features — braces, joints, weld caps, repairs — contaminate fitted geometry and produce misleading results if they are included uncritically. Getting the fits right is analytical work, not a processing step, and it is where the value of the exercise is either created or lost.
Two constraints worth knowing early
Laser scanning is line-of-sight. Insulation and cladding obscure the shell beneath, so a lagged vessel cannot be assessed through its cladding. Where the shell is the subject, either the cladding comes off or the scope changes.
Internal scanning needs safe access and, for vessels and tanks, normally an empty asset. On live plant this is the binding constraint on when the work can happen, and it usually means tying the survey to a planned shutdown rather than treating it as a standalone job.
Neither prevents useful work. Both are much cheaper to design around at scoping than to discover on site.
If you have no baseline
The honest answer is that you cannot recover history. What you can do is establish the datum now, so that the same question in two years has a real answer.
A first survey still has immediate value — measured as-built condition, out-of-roundness and alignment against design intent, and a record that can be compared against drawings. But its main value is as the reference for the next one, and that value only accrues if the control is done properly the first time.
Establishing a baseline
If you have an asset you are uncertain about, the first conversation is about what you suspect is moving and what decision the answer feeds. That determines whether this is a baseline exercise, a comparison against existing data, or a condition survey against design intent.