Products · Industrial Multirotor
Thermal Imaging Payload Configuration for Inspection
A thermal payload is only as useful as its calibration and its recording discipline. Two things decide whether the images can be compared later: emissivity assumptions and the reference temperature at capture.
Overview
- Record ambient temperature, humidity and time with every sortie — a thermal image without context is not evidence.
- Different surface finishes have different emissivity. Metal fittings and painted panels are not directly comparable.
- Set the palette and temperature range before the flight, not after. Auto-scaling hides the absolute values.
- Keep the same camera, lens and altitude across a survey series if you intend to compare the results.
What thermal is good at, and what it is not
Thermal imaging is good at finding temperature differences that are invisible to the eye: an overheating connector, a delaminated panel, a wet patch in insulation, a blocked cell in an array. It is not a fault diagnosis. A hot spot tells you something is dissipating more than its neighbours; it does not tell you why, or whether it matters yet.
Treating thermal output as a diagnosis is the single most common failure in these programmes. It produces confident-looking reports that engineers then have to re-check, and after two cycles the programme loses trust and budget.
Conditions decide whether the reading means anything
A thermal reading is a comparison against assumptions: emissivity, reflected temperature, distance, atmospheric transmission, and the load on the asset at the moment of capture. Change any of them and the number moves — sometimes enough to turn a marginal finding into a false alarm, or a real one into a miss.
So the configuration has to capture conditions with the image. We record asset identifier, position, time, ambient conditions, load where available, and the emissivity setting, and we state that confirmation is a separate inspection step. Where a figure matters as an acceptance criterion, we issue the test conditions for the operating case rather than a bare number.
Radiometric versus visual thermal
If you need to compare temperatures, you need radiometric capture with per-pixel temperature data and a known emissivity setting. If you only need to see that something is hotter than its surroundings, a non-radiometric feed is sufficient and considerably cheaper. Many programmes buy radiometric capability they never use because nobody wrote down which of the two they needed.
Decide that before procurement. It changes the payload, the data volume, the file formats your downstream tooling must accept, and the skill needed to read the output.
Where the review step belongs
Automated thermal detection produces candidates, not conclusions. False positives come from reflection, solar loading, soiling, wet surfaces and edge effects — all of which look exactly like faults to a detector that has no context.
Place review where the reviewer has both the thermal frame and the visible frame of the same point, plus the capture conditions. That is the difference between a list that gets acted on and a list that gets ignored. See M-15 for the paired-payload configuration this implies.
Related pages
- M-30 heavy-payload long-endurance multirotorView
- M-15 with thermal imaging optionView
- M-10 compact inspection platformView
- Intelligent Payloads and AccessoriesView
More in this area
- How to Choose an Industrial Multirotor PlatformRead
- Multirotor Configuration for Power Facility InspectionRead
- Multirotor Configuration for Solar PV InspectionRead
- Multirotor Configuration for Building Facade InspectionRead
- Multirotor Configuration for Bridge InspectionRead
- Thermal Imaging Payload Configuration for InspectionThis page
- Emergency Lighting and Speaker Payload ConfigurationRead
- Multi-Payload Platform ConfigurationRead
- Endurance and Payload Trade-offs for MultirotorsRead
- Multirotor Customization Requirement ChecklistRead