More filters
The same selection, counted. Set filters on the Search tab and these follow. Click any bar to filter on it.
Reports per month
Aircraft system
Airline
Model
Part
What was found
What the crew did
How it was found
Words that stand out here
The vocabulary of this selection against all 170,201 reports. Useful for finding the term the trade actually uses, which is rarely the one you searched for.
One aircraft at a time. Type a tail number, or click one anywhere in the tool.
Most written-up aircraft in the current selection
Where to start. Both tables compare the last 90 days with what came before.
Aircraft with a sudden cluster
What is rising
One part number, failing the same way, on many aircraft and at more than one airline. That is a fleet problem, not an incident.
One airline, one type
A count of reports means little on its own. What matters is how many separate aircraft it touched. Ten write-ups on one airframe is a bad machine. Ten write-ups across ten airframes is a fleet problem.
Defects that are new
A defect that is new tells you more than one that is merely common. These are parts and systems being written up now that barely appeared before, so the pattern is still forming. Small numbers, read them as a tip rather than a finding.
Same airline, same system, same day
When one airline writes up the same system on several different aircraft on a single day, the cause is rarely the aircraft. It is a batch of parts, a procedure, a supplier or a shared inspection. This is the pattern that turns a maintenance note into a story.
Corrosion and cracks
The slow story. Corrosion is graded by the FAA. Level 2 means it went past what the manufacturer allows and needed repair. Level 3 is an urgent airworthiness concern: the operator must notify the regulator within three days and act across the whole fleet, not just on that aircraft. Level 1 stays within limits and is not reportable, which is why you never see it here.
Do old airframes break differently?
Hours and cycles are two different kinds of old. A short-haul aircraft piles up cycles fast and hours slowly, and since every flight pressurises and depressurises the hull, it is cycles that drive cracking. Hours wear out the things that simply run.
Engines
Handle with care. Only about three in a hundred reports name the engine, so this is a count inside a small subset, not a failure rate per engine in service. Use it to find cases to check, never to rank manufacturers.
What the crew actually had to do
Not what broke, but what it forced.
How was it found?
A crack found by eddy current or X-ray was invisible from the outside. One found by eye was not. Splitting a system's findings by method tells you whether trouble is being caught by instruments before anyone could see it, or noticed only once it showed. Not every report falls either way: functional checks and unrecorded methods make up a large share, and they are listed too.
This follows whatever you have filtered on the Search tab.
Two airlines or models side by side. These are raw counts: a bigger airline files more reports, so this is a starting point, not a score.
Everything in this tool, in plain words. The shorthand comes from maintenance logs; the codes come from the FAA's own tables.
What you are reading
FAA codes
Where this comes from
Operators and mechanics must report failures, malfunctions and defects to the FAA. Those Service Difficulty Reports are public. This holds every report the FAA published for , in total, taken from the yearly files at external.apic4e.faa.gov/sdrs/retrieve/SDR-YYYY.csv. There is no FAA API; the files are downloaded whole and rebuilt here.
What it cannot tell you
This is not an accident database and not a safety ranking. A high count can mean an old fleet, a thorough maintenance department, or simply a large airline. Nothing here is divided by fleet size or flying hours, because the FAA file does not contain either.
A write-up is a defect that maintenance caught. That is usually the system working, not failing.
About the codes
Every coded field is decoded from the FAA's own lookup tables, which the FAA publishes as a zip file on the SDRS front page. This site carries the edition dated July 2026, shipped with the code so you can check it. Run against all 170,201 reports, those tables resolve every code that actually occurs, with none left over.
An earlier version of this page decoded the letters from a third-party copy of the tables,
checked against three codes in the FAA's field instructions. That copy was wrong in at least one
place: it had R as engine flameout, where the FAA's own table has R as
a partial loss of rpm or power and X as flameout. Checking three cells said nothing
about the rest. The tables now come from the FAA directly.
What this tool promises
Cold start to a quotable sentence on your clipboard in six gestures: click a starter question, click the operator in the WHO rail, drag a range across WHEN, click Case sheet on a row, read the quote, click Copy all three. Before this rewrite the same task took about eighteen, ten of them keystrokes inside two date fields. Nothing ships here that lengthens it.
One thing here is not decoded from the FAA's own tables: the corrosion grades, which are not in the FAA zip. Those definitions come from FAA Order 8300.12 and EASA IP 119 and are labelled as such wherever they appear. The four-letter operator designator is expanded from a different FAA source, the Air Carrier/Operator cross-reference, December 2006 edition, which is the most recent one obtainable. 183 of the 309 designators in this data resolve, covering 98.7% of all reports. The rest are shown exactly as filed. Any name can be stale where a carrier has merged, been renamed, or kept filing under a legacy code, so check current ownership before you publish.
Before you publish
Open the original report through the FAA's own query page, check the tail number in the FAA registry, and put the finding to the airline. Read the write-up itself rather than the count: the mechanic's sentence is the evidence.