What should a maintenance team record during a thermal scan?

5 min read · Updated 30 July 2026

There is a simple test for a scan record: could a colleague who was not there — or you, in two years — act on it without phoning anybody? Repair from it, re-scan against it, defend it to an auditor. Everything below exists to pass that test, and none of it takes longer than the walking between panels.

Per panel, whether or not anything is wrong.

The image pair — thermal and ordinary photo, same view, seconds apart. The load — what was running, and the per-phase current where the panel metering shows it; a reading without its load is unverifiable, and "unverifiable" is what an insurer's assessor calls "worthless" politely. The identity — panel name or number exactly as the site knows it, not "the grey board near the compressor." Healthy panels get this too: today's unremarkable image is next year's baseline, and a scan that only photographs problems has no way to prove anything else was checked.

Per finding, three more lines.

The difference — not just the temperature, but what it was compared against: the same phase on the neighbouring circuit, the adjacent breaker, ambient. A number with a named reference is a measurement; alone it is trivia. The call — monitor, plan the repair, or act now, using whatever severity bands your reports already use, stated so the next reader applies the same logic. The location, findable — labelled to the terminal, so the electrician repairs the right joint and the verification re-scan afterwards targets the same point. A finding that cannot be found again cannot be closed.

The folder is half the value.

One folder per survey, named by date. Same structure every time, so the 2026 folder sits beside 2025's and the same panel's images line up across years. This is the unglamorous machinery behind every trend finding — "2°C over reference last year, 9°C now" is only possible when both readings can be found, compared like-for-like, and trusted. It is also, frankly, what separates an in-house scanning habit that professional surveyors respect from a camera roll of orphaned rainbow images. The camera makes the pictures; the discipline makes the evidence.