Trust & transparency

Data

iamSafer enriches your logbook to power its stats. It happens in stages, and every stage is best-effort — some of your flying carries all of it, some none. Rather than assume completeness, the Data coverage section of your stats shows exactly how far each stage reaches for your logbook.

The stages

• GPS tracks. We match your logged flights to recorded GPS flight tracks. Not every flight can be matched — older flights, or flights a provider never recorded — so some have no track at all. A single flight can also carry more than one track.
• Flight operations. From a matched track we detect each takeoff and landing. How we detect them ↓
• Weather. For each detected operation we capture the weather observed at that moment. How we match it ↓

Because each stage only reaches a subset of the one before it, the coverage figures step down — that drop-off is expected, and showing it is the point: a small, favourable sample should never be mistaken for your whole logbook.

Flight operations

An operation is a single takeoff or a single landing — the unit most of the weather and currency figures are counted in.

How we detect them

We derive operations from your matched GPS track — identifying the moments the aircraft left and returned to the ground, and the runway used where it can be resolved. Detection is best-effort and depends on the track’s quality and coverage: a flight with no matched track has no detected operations, and a partial track may miss some. These detected operations are what weather is then attached to, which is why weather coverage can never exceed operation coverage.

Detected operations are distinct from the takeoff and landing counts you enter by hand in a logbook entry — those are always kept as you recorded them; detection only adds the ones we can find on a track.

Weather flown

The weather figures in your logbook stats describe the conditions you’ve actually operated in, reconstructed from the weather observed at each of your takeoffs and landings.

Where the data comes from

When one of your flights is matched to a flight track, we identify each takeoff and landing on it and pair that moment with the nearest applicable weather observation (a METAR) for the airport. Wind direction and speed, gusts, the resulting crosswind component, and the reported flight category (VFR / MVFR / IFR / LIFR) come from that observation. Only observations we could actually resolve are used; when we looked and found nothing usable, that operation contributes nothing rather than a guessed value.

The crosswind is a computed component, not a reported value: it needs both the runway used and a definite wind direction to work out. An operation flown on a runway we couldn’t resolve — or in a variable (“VRB”) wind with no single direction — still records its wind, but no crosswind. So a given aircraft can show a wind figure with no crosswind beside it: the operation counts toward the wind distribution but not the crosswind one.

The wind & crosswind bands

Each band shows the middle of your distribution — the range from the 10th to the 90th percentile of every observation, with a tick at the median (50th percentile). It answers “what do I routinely operate in?”, not “what’s the worst I’ve seen?”

The bar is drawn on a fixed scale — wind 0–25 kt, crosswind 0–15 kt — so the shape is comparable between pilots and over time, and one unusual flight can’t stretch it. The numbers printed on each band are always your true percentiles; a value beyond the end of the scale is pinned to the edge of the bar but still reported accurately in the label.

A band needs at least five observations to be drawn. Below that we show the median tick alone, without the shaded 10th–90th range — a spread inferred from two or three flights would imply a precision the sample can’t support. Wind and crosswind are counted separately, so one may have enough observations for a full band while the other shows only its median.

Coverage — “N of M operations”

Weather is matched on a best-effort basis. The coverage figure compares the operations we captured a weather observation for (N) against every takeoff and landing in your logbook (M). Many operations have no weather attached — older imports, flights we couldn’t match to a track, or moments with no usable nearby observation — so coverage is frequently a small fraction of your total flying. That is expected, and it’s why we show the figure: so a small, favourable sample is never mistaken for the whole picture.

Why distributions, and not records

We deliberately do not show a “maximum crosswind” or any similar personal record. A single extreme is one data point, stripped of the context that gives it meaning — the aircraft, your currency at the time, whether it was flown dual, and how it actually turned out. Reading a lifetime maximum as a personal capability (“I’ve handled it once, so I can handle it”) is exactly the reasoning that gets pilots into trouble. Distributions show what you routinely fly in without putting an outlier on a pedestal.