Insights / Inspection

Facade inspection without scaffolding — where AI screening fits

Most of a facade survey budget is spent finding defects, not assessing them. Narrow the search and the detailed survey gets cheaper. This is about how to narrow it.

Insights / Facade inspection without scaffolding — where AI screening fits

The expensive part is the search

What inflates a facade survey quote is usually not the detailed assessment. It is the step before it. Because nobody knows yet which areas have deteriorated, scaffolding goes up around the whole building and the whole surface gets tested. If the damage is concentrated in thirty per cent of the area, you still pay for one hundred per cent of it.

Put the other way round: if you already know which areas are suspect, the detailed survey only has to cover those. Spend on the first pass to shrink the second. Aircraft maintenance works the same way — you do not strip the airframe, you follow the system that showed an indication.

Three methods, three characters

The ways of finding facade deterioration differ in what they trade away — accuracy, cost, or coverage.

MethodGood atPoor at
Visual check from ground levelCosts nothing. Starts immediatelyCannot see upper floors or shaded elevations. Judgement varies by inspector
Hammer (percussion) testingReliably finds delamination. Accepted for statutory surveysNeeds scaffolding or specialist access. Heavy on cost and schedule
AI assessment from photographsCovers large areas quickly. Leaves a record you can compare over timeCannot see subsurface delamination. A surface-level first pass only

These do not compete. AI assessment is not a substitute for percussion testing — it works when you treat it as the tool that decides where percussion testing should go.

Starting from photographs you already have

If site photographs already exist, that is the fastest way in. Ordinary smartphone photographs are usually enough: what matters is that the area of interest fills the frame and is adequately lit. Cracking, spalling, rust and water leak staining are all visible at the surface, which is exactly what a photograph captures.

The practical output to aim for is not a list of defects but a ranked one. A report that says "eleven findings" helps nobody; a report that says which three need attention this quarter changes what you do on Monday. Severity bands and a numeric urgency score are what make a survey actionable.

The inspection tooling we build for this sits in Airffic World, which takes uploaded photographs and returns findings with severity and an urgency score.

Where it belongs in a renovation plan

For a large-scale repair programme on an apartment building, the first pass largely determines how accurate the plan is. When the estimated scope diverges from what is actually there, extra work appears after contracts are signed, and the owners' association has to rebuild its consensus from scratch.

Getting the distribution of deterioration before the plan is written keeps the budget anchored to the building rather than to an assumption.

The limit worth stating plainly

AI assessment from photographs is a first-pass screen. It does not replace statutory inspection or a detailed survey by a qualified engineer. Subsurface delamination and structural damage are not visible at the surface, and no amount of image processing invents that information. The workable pattern is to send a qualified professional to the areas the screen flags as high severity, in that order.