How accurately Hobbs reads a paper logbook

We measure the scanner against hand-verified logbook pages every time it changes, and we publish the result, including what it gets wrong.

The numbers

  • 99.5% of flights found on photographed pages.
  • 99.2% of hours exact to the tenth: total, day, night, cross-country, instrument and simulator time.
  • 100% of take-off and landing counts written on the page read correctly.
  • 97.4% of identity fields (date, aircraft type, registration, names, route) correct before the pilot touches anything.
  • 99.98% agreement on the grand total: the hours Hobbs adds up match the totals printed on the pages to within a tenth of an hour.
  • About 28 seconds a page (the slowest page took 37).

Measured 12 November 2026 · Hobbs scanner as deployed that day

What the pilot still checks

Four in five flights came through with nothing to correct at all. The mistakes that remain are mostly aircraft registrations (90% read right first time: a "Q" that looks like an "O", or a tail written in a hurry) and names. Those are exactly the fields Hobbs is built to catch:

What this test is, and isn't

The pages are real logbook pages (a mix of dual instruction, solo, night, instrument and cross-country flying) photographed with a phone in ordinary light. It is not a guarantee for every handwriting or every logbook layout. Your own pages may read better or worse. That's why every new account can scan 20 flights (about two pages) free, and why nothing is committed until you've seen it.

The test runs the same code the app runs, every time the scanner changes. It scores the scanner only on what is actually written on the paper, so the numbers mean what they say.

Try it on your own handwriting

Scan 20 flights free