Almost every wasted thermal sortie we hear about was preventable on the ground. The checklist below is built from the failures operators actually report, not from a generic template — each item is there because skipping it has cost somebody a flight.
Key takeaways
- Almost every wasted thermal sortie was preventable on the ground — the failures cluster in a small number of repeatable checks.
- Four groups cover it: mechanical and installation, electrical and interface, imaging and radiometric, and mission-specific settings.
- The radiometric settings — emissivity, reflected temperature, gain mode — are the items most often skipped and the ones that most often invalidate the data.
- A checklist only works if it is enforced procedurally; make it a sign-off, not a habit.
On this page
Mechanical and installation
- 1. Gimbal free through full travel. Power on and confirm the payload sweeps its full range without binding on cabling or mount.
- 2. Transport lock removed. The single most common ground fault, and it produces motor errors that look alarming.
- 3. Isolators intact and correctly loaded. Check for cracking, hardening and correct static sag. Cold weather stiffens elastomer.
- 4. Cable strain relief. No tension on the payload connector through the full gimbal travel range.
- 5. Mount fasteners torqued. Vibration loosens them; a loose mount produces image problems that look like gimbal faults.
- 6. Lens and window clean. A fingerprint on a germanium window is a permanent soft patch in every frame of the sortie.
Electrical and interface
- 7. Connectors fully mated and sealed. Ingress almost always enters through a partly mated connector, not through the housing.
- 8. Supply voltage within range under load. Check with motors spinning, not at idle.
- 9. Control link verified from the actual autopilot. Not from the bench tool you used last week.
- 10. Video downlink at operating range. Confirm at distance, not next to the aircraft.
- 11. Firmware baseline matches the fleet record. See firmware architecture for why mixed baselines cause unreproducible faults.
- 12. Storage card present, formatted, with capacity for the planned sortie.
Imaging and radiometric
- 13. Warm-up completed. 5–15 minutes powered before any measurement. Take-off is not the start of the warm-up.
- 14. FFC performed and image uniform. Residual pattern noise means the correction did not take.
- 15. Both channels checked. Thermal and visible confirmed live, and the overlay checked while slewing.
- 16. Zoom and focus through full range. Hunting at the long end shows up here, not in the air.
- 17. Emissivity set for the target material. Not left at the default from the last job.
- 18. Reflected apparent temperature entered. Especially for shiny surfaces under open sky.
- 19. Distance set or laser ranging confirmed working. Required beyond about 100 m for accurate radiometry.
- 20. Palette and level/span appropriate for the expected scene range.
Mission-specific
- 21. Conditions recorded. Ambient temperature, wind, cloud, and — for electrical work — equipment load. Without these the findings are not defensible; see report deliverables.
- 22. Timing appropriate to the task. Pre-dawn for buried services, mid-afternoon for delamination, night for search, high solar load for PV. Flying the right mission at the wrong hour is the most expensive checklist item to get wrong.
Three items worth calling out
| Item | What happens if skipped | Cost |
|---|---|---|
| Warm-up | Radiometric readings drift several kelvin | Whole survey unusable for measurement |
| Emissivity set per job | Systematic under-read on metal | Findings understated, some missed entirely |
| Conditions recorded | Findings cannot be reproduced or classified | Report rejected at review |
| Isolator condition | Jello and rolling-shutter skew | Imagery unusable at high zoom |
| Overlay check while slewing | Timing or boresight error undetected | Fusion and geolocation wrong all sortie |
| Timing to task | No thermal contrast to find | Flight completes successfully and finds nothing |
Related reading
Operations: seven thermal imaging mistakes, gimbal maintenance and cold-weather operations.
Integration: payload integration checklist and vibration and damping design.
Making the checklist stick
A checklist that lives in a folder is a checklist that gets skipped on the third busy morning. Three things determine whether it survives contact with an operational tempo.
Split it by phase. Items 1–6 belong to assembly, 7–12 to power-on, 13–20 to the payload warm-up window, and 21–22 to the mission brief. Grouped that way each block takes under two minutes and fits naturally into what the crew is already doing, instead of being one long interruption.
Make it challenge-and-response where there are two crew. One reads, one acts and confirms. Single-pilot self-checking is where items get silently skipped, and the industry has known this for a very long time.
Finally, record it. A dated, signed checklist takes ten seconds and is the difference between “we always check that” and evidence that you did — which matters on the day a client disputes a finding or an insurer asks a question.
- Group the checklist by phase so no single block exceeds two minutes.
- Use challenge-and-response with two crew; assume solo self-checking will drift.
- Complete items 13–20 during the warm-up window rather than adding them to it.
- Record and retain the completed checklist with the flight log.
- Review the checklist quarterly and add any failure the team actually experienced — that is how it stays honest.
FAQ
How long does the payload need to warm up?
5–15 minutes powered before the first measurement. Uncooled detectors drift while the substrate temperature stabilises, and readings taken during climb-out are the least reliable of the sortie. Power the payload while you complete the rest of the pre-flight.
Which checklist item is skipped most often?
Recording conditions — ambient temperature, wind and equipment load. It feels like paperwork rather than a check, but findings without it cannot be classified or reproduced, and the report gets rejected at review.
Do I need the full checklist for a search flight?
No. Radiometric items 17–19 do not apply to search work. Items 1–16 and 22 do, and item 22 — timing to task — matters more on search than anywhere else.