Thermal surveys map trapped moisture across a large roof in one flight — but every warm patch is a hypothesis until someone verifies it on foot. Here is how to buy one properly.
A thermal camera does not see water. It reads infrared radiation coming off a surface and turns it into a temperature map. What makes the technique useful on a roof is a property of water rather than a property of the camera: water stores a great deal of heat for its volume, so a wet patch of insulation, screed or board absorbs solar heat all day and then gives it back slowly through the evening, while the dry material around it cools quickly. Fly at the right moment and that difference shows up as a warm blotch with a soft, irregular edge sitting inside a cool field.
Two things follow from that, and they set the boundaries of the whole service. First, the camera is reading the top surface — it infers what is underneath from how that surface behaves, which is why the technique works on an exposed membrane and gets progressively less reliable as you bury it under paving, ballast, planting or panels. Second, anything else that makes one part of the roof warmer than another produces the same picture: a plant room below, a heated production line, a lit warehouse bay, a duct running under the deck, even a different colour of patch repair. A thermal image is evidence of a temperature difference and nothing more. Interpreting it is the job; flying the drone is the easy part.
This is the single most important sentence on this page, and it is the one that separates a survey worth paying for from an expensive picture. Every anomaly on a thermal map is a hypothesis until it is verified on foot. A responsible operator marks the anomaly, describes it, and then either ground-truths it there and then or hands you a register of locations to be checked before anyone prices a repair. An operator who hands you a colourful PDF, calls the warm areas “wet” and invites you to tender on it has skipped the part that carries the risk.
The cost of skipping it is asymmetric. A false positive means you strip and replace a dry bay — wasted money, but visible. A false negative means a genuinely wet zone stays in the building because it sat under a panel, under standing water or over a warm room, and you have now signed off a roof that is still failing. Ground-truthing catches both. It is cheap relative to the repair and it is the difference between a moisture map and a guess.
The contrast the camera depends on only exists for part of the day. Through the afternoon the whole roof is hot and the wet areas are hidden inside that heat. After sunset the dry build-up sheds its heat quickly while the wet material lags, and the pattern emerges. That is the window a survey should be booked into, and it is why a thermal flight is usually an evening job rather than a convenient mid-morning one.
Weather ruins it just as effectively as timing. Rain during or shortly before the flight wets everything and flattens the contrast the whole method relies on. Wind moves heat off the surface unevenly. Heavy cloud through the day means the roof never took on enough solar gain to have anything to release. In a Klang Valley monsoon month a survey may simply have to wait for a dry run of days, and any operator who promises a fixed date regardless of weather is promising something they cannot control. Build that into the programme: agree the standby arrangement and who bears the cost of an aborted flight before the job starts, not on the night.
Used inside its limits, the technique is very good at a specific class of problem: large, flat, exposed roofs where the leak has already got into the build-up and nobody can tell how far it has travelled.
| What it is good at | Why the method suits it | What you do with the result |
|---|---|---|
| Mapping the extent of trapped moisture | Wet build-up holds heat and reads warm against dry material around it | Draw the zone boundary for a repair-versus-replace decision |
| Large factory, warehouse and podium roofs | Covers in one flight what takes days to walk safely | Prioritise which bays are tendered first |
| Roofs that cannot be walked safely | Fragile skylight sheets and corroded decks are dangerous underfoot | Survey without putting anyone on the roof |
| Deciding between overlay and strip-and-replace | Shows whether wet material is isolated or widespread | Avoid coating over a saturated build-up |
| Recording a baseline after a re-waterproofing job | A dry roof photographed thermally is a reference for later | Compare against the next survey to see what changed |
| Settling a “where is it actually coming from” argument | Puts a mapped pattern on the table instead of two opinions | Scope the investigation instead of arguing about it |
Send us a photo of the job on WhatsApp — we reply with an instant quote.
💬 Get Your Instant QuoteThe common thread is extent. If you already know a roof leaks and only need to know how much of it is compromised, this is the tool. Our factory and warehouse roof waterproofing guide covers what the repair itself looks like at that scale.
The honest list matters more than the sales list, because most disappointed clients were sold the technique for a job it was never going to do.
| Situation | Why thermal struggles | Better first step |
|---|---|---|
| Standing water or ponding on the roof | The pond has its own thermal mass and masks whatever is beneath it | Clear the outlets, let it drain, resurvey |
| Under solar panels and their mounting rails | The camera sees the panel, not the roof under it | Physical inspection at bracket penetrations |
| Under thick paving, ballast or planting | Too much material between the wet layer and the surface | Test-open at the low points and outlets |
| Over plant rooms, ovens, lit bays or hot ducts | Heat from below reads exactly like trapped moisture | Overlay the building services plan before interpreting |
| A single active drip in a house or shoplot | The roof is small enough to inspect properly on foot | Leak detection with meter and dye tests |
| Pipework and concealed plumbing leaks | The leak is inside the structure, not in the roof build-up | Concealed pipe leak detection |
| Which layer failed, and why | A temperature map shows a symptom, not a defect | Test cut and a core through the build-up |
Note the last row especially. A thermal survey will not tell you whether a lap failed, a fastener washer perished or a flashing detail was wrong. Those are found by opening the roof at the anomaly — see our metal roof leak repair guide for the failure points a test cut is looking for.
These are complementary rather than competing, and the sensible sequence uses both. A moisture meter measures the material directly at the point you press it against, which is authoritative but slow: on a hundred-thousand-square-foot factory roof you would be there for days and you would still only sample a grid. A drone thermal flight covers the same roof in an evening but only produces a pattern. So fly first to decide where to press the meter, and press the meter to decide whether the pattern is real. Neither half is optional.
There is also a safety argument that carries real weight on ageing industrial roofs. Brittle translucent sheets, corroded purlins and rusted fixings make walking a large roof genuinely hazardous, and the fewer trips across it before the scope is settled, the better. A flight narrows a whole roof down to a handful of places anyone actually has to stand.
This is the workflow that turns a survey into something you can tender. Each anomaly should travel through the same four steps, and the result of each step should be written down against the anomaly reference.
ClickBina prices this verification stage as a leak detection and assessment visit (RM300 – RM800 per visit), and the fee is waived if you proceed with the repair. It is the same discipline described in our water leak detection service guide, applied to a roof instead of a bathroom.
Ask for the report specification before you accept a quote, because the report is the product. At minimum it should carry: the date, time and weather conditions of the flight and why that window was chosen; a plan of the roof with every anomaly numbered and positioned; the thermal image and the matching daylight image for each anomaly; the ground-truth result recorded against each one; a clear statement of the areas that could not be surveyed and why; and a recommended next step per zone rather than a single verdict for the whole roof. Ask for the raw imagery to be handed over as well, so a second opinion is possible without a second flight.
Two lines are worth insisting on and are often missing. One: a stated confidence level per anomaly, so you can see which findings the operator would defend and which they would not. Two: a note of anything that limited the survey on the night — wind, cloud cover, a wet afternoon, a section shut off for operational reasons. A report with no limitations section is a report that has not been written honestly.
The point of the exercise is a decision, and there are only a few decisions available. Where the wet zones are isolated and the surrounding build-up is sound, you cut out and make good locally, then treat the roof normally. Where wet zones are widespread, coating over the top traps the water you have just paid to find, and the honest answer is strip and replace those bays. Where the survey is inconclusive over a large area, the next spend is investigation, not repair.
Whatever the outcome, the survey should feed a scope of works with quantities attached — square feet of build-up to be removed, number of test cuts to be made good, linear metres of flashing and lap to be re-detailed, and the coating or membrane specified for the balance of the roof. That is what turns three quotes into a like-for-like comparison, and it is the same principle set out in our waterproofing quotation guide. Roof system rates for the repair itself sit in the usual bands (market estimate 2026, indicative Klang Valley):
| System | Market range (2026, Klang Valley) | What it is for |
|---|---|---|
| Acrylic roof coating | RM4 – RM10 per sq ft | UV topcoat for tiled and metal roofs |
| Cementitious slurry | RM6 – RM10 per sq ft | rigid; best on sound concrete |
| Torch-on bituminous membrane | RM8 – RM15 per sq ft | flat roofs and podium decks |
| Liquid PU membrane | RM10 – RM18 per sq ft | seamless, the common flat-roof choice |
| Roof waterproofing (all systems) | RM8 – RM20 per sq ft | the whole-roof band across systems |
Those are the rates that get applied to the areas the survey defines. Whole-roof context is in roof waterproofing cost, and the per-square-foot logic across systems is in waterproofing cost per square foot.
A flight over an occupied industrial site is a coordination job before it is a technical one. Before the survey date, settle in writing: who is on site that evening and where they will be; whether the roof needs to be cleared of contractors; where the pilot will launch and land; how the flight interacts with any operations below; and whether any part of the site is off limits. Ask the operator to provide their approvals and liability cover in writing, and keep a copy on file with the report — that is a procurement step, and it is easier to ask for before the flight than after an incident.
The other half is what happens the moment the map exists. Ground-truthing puts people on a roof that may be fragile, at night or the following morning, so the same access controls, fall protection and permit-to-work arrangements you would use for any roof work apply to the verification visit too. Budget for that access rather than discovering it when the first test cut is due.
It earns its keep when the roof is large, flat and exposed, when the leak is diffuse rather than a single drip, when the building is operating below and you want to minimise trips across the deck, when the decision at stake is repair-versus-replace over a big area, and when a committee or a head office needs documented evidence before releasing money. Factories, warehouses, podium decks, school and institutional blocks and large flat-roofed commercial buildings are the natural fit.
It is the wrong first spend on a terrace house, a shoplot or a condo unit with one active drip — those are solved faster and cheaper by a technician with a meter, a dye test and a pressure test. It is also a poor fit on a pitched tiled roof, where the water path is under the tiles rather than through a saturated build-up, and on any roof that is currently ponding or fully covered by panels, paving or planting. In those cases the money is better spent on detection or on opening the roof at the obvious suspects first.
A drone thermal survey is priced per project, not per square foot, because the drivers are roof area, the number of separate roof levels, access and airspace constraints, how much of the roof is obstructed, whether the flight has to be scheduled outside operating hours, and how much ground-truthing is included. Ask every quote to state plainly whether verification and the written report are inside the price or extra — that is where two survey quotes usually differ, and a cheap flight with the report priced separately is not the bargain it looks like.
Verification and the repair itself are priced on ClickBina’s normal terms (market estimate 2026, indicative Klang Valley for the market rows):
| ClickBina service | Price | What it covers |
|---|---|---|
| Leak detection & assessment | RM300 – RM800 per visit | waived if you proceed with the repair |
| Tiled / metal roof repairs | RM500 – RM3,000 per job | spot repairs at verified anomalies, before any full recoat |
| Roof waterproofing (all systems) | RM8 – RM20 per sq ft | full system across the zones the survey defines |
Read the second and third rows as the two outcomes a survey produces: a handful of located repairs, or a system across defined bays. The survey is what tells you which conversation you are in, and that is the whole return on it — see waterproofing scams for the pattern where a full-roof price is quoted without anyone establishing extent first.
Treat repeat surveys as event-driven rather than calendar-driven. The occasions that justify another flight: immediately after a major re-waterproofing job, to bank a dry baseline while you know the roof is sound; before the monsoon on a roof with a known history; after a storm that lifted or damaged sheets; when internal staining reappears in a bay that was previously signed off; and before a large capital decision such as a re-roof tender or a building handover, where a documented moisture map changes what the tender asks for.
Between surveys, the cheap work is the ordinary work: clear the gutters and outlets, check the flashings and laps, and log every internal stain with a date and a location so the next survey has something to correlate against. Our scheduled roof maintenance and monsoon roof preparation checklist guides cover that routine; gutter and downpipe repair covers the drainage side that causes most of the ponding a survey has to work around.
ClickBina works the roof end of this problem across Kuala Lumpur and Selangor: we verify what a survey has flagged, open the roof where the evidence points, and price the repair against a scope with quantities rather than a colour picture. Where a thermal survey is the right first step on a large roof we will say so, and where a meter, a dye test and an hour on the roof would settle it faster and cheaper we will say that instead. Detection and assessment is RM300 – RM800 per visit, waived if you proceed with the repair. Send us the roof area, the building type and photographs of the internal staining on WhatsApp and we will tell you which route makes sense. Related reading: waterproofing services, expansion joint leak repair and podium deck waterproofing.
Tell us what you need — we reply within the hour.