himaxym/faa-aviation-safety-rollups

FAA wildlife strikes, laser incidents and drone sightings — analysis-ready rollups

Three United States FAA safety datasets, cleaned and rolled up into small tabular files you can load without touching the source archives.

This is not a copy of the FAA's raw releases. Those are already public and large. What is here is the part that does not exist upstream in this form: stable slugs, consistent naming, and per-airport / per-species / per-state / per-aircraft / per-year rollups computed once over the full record set.

Files

Wildlife strikes — total counts strikes, damaging counts those that caused aircraft damage.

• —wildlife_strikes_by_year.csv — 37 rows: year, total, damaging
• —wildlife_strikes_by_airport.csv — 454 rows: slug, name, state, total, damaging
• —wildlife_strikes_by_species.csv — 288 rows: slug, name, total, damaging
• —wildlife_strikes_by_aircraft.csv — 223 rows: slug, name, total, damaging

Laser incidents — lasers aimed at aircraft, as reported by flight crews.

• —laser_incidents_by_year.csv — 5 rows: year, total, injuries
• —laser_incidents_by_state.csv — 56 rows: slug, name, total, injuries
• —laser_incidents_by_airport.csv — 183 rows
• —laser_incidents_by_city.csv — 96 rows
• —laser_incidents_by_aircraft.csv — 107 rows

Drone sightings — UAS sightings reported by pilots and air traffic controllers.

• —drone_sightings_by_year.csv — 8 rows
• —drone_sightings_by_state.csv — 60 rows
• —drone_sightings_by_city.csv — 80 rows

manifest.json carries the export date and the headline totals for each source.

Loading

from datasets import load_dataset

by_species = load_dataset("himaxym/faa-aviation-safety-rollups", "wildlife_by_species")
by_year    = load_dataset("himaxym/faa-aviation-safety-rollups", "wildlife_by_year")

Things worth knowing before you use it

• —The year columns are counts of reports, not risk. Wildlife strike reporting is voluntary and reporting rates have risen sharply over the period — the climb from 2,120 strikes in 1990 to 24,459 in 2025 is substantially a reporting artefact, not a fourteen-fold rise in bird strikes. Treat any year-over-year trend as a statement about the reporting system unless you control for it.
• —2026 is a partial year in all three datasets.
• —The laser series starts in 2021 because that is where the FAA's current public release starts, not because lasers began then.
• —Airport and city rows are only those the FAA named. Reports with an unknown or blank location are counted in the totals but appear in no location rollup, so location rollups do not sum to the dataset total.
• —Species names are as recorded by the reporter, including generic groups ("Gulls", "Sparrows") alongside precise identifications; they are not a taxonomy and should not be joined to one without cleaning.
• —No coordinates are included. The upstream geocoding is airport-anchored and derived from an ODbL source, so shipping it under a public-domain banner would misstate both its licence and its precision.

Provenance and licence

Underlying data is United States Government public domain (17 U.S.C. §105), from:

• —FAA National Wildlife Strike Database — <https://wildlife.faa.gov>
• —FAA Reported Laser Incidents — <https://www.faa.gov/aircraft/safety/report/laserinfo>
• —FAA UAS Sighting Reports — <https://www.faa.gov/uas/resources/public_records>

The rollups, slugs and normalisation in this repository are released under CC BY 4.0 by FlightFinder. Attribution is appreciated and required for the derived layer; the underlying FAA records carry no such requirement.

Live and complete data

These files are a snapshot of aggregates. The full occurrence-level data, kept current, is available as a free JSON API and as an MCP server:

• —API docs and keys — <https://himaxym.com/developers>
• —MCP server — https://himaxym.com/mcp (registered as com.himaxym/flightfinder-aviation-safety)

Both answer without a key at the keyless tier:

curl 'https://himaxym.com/api/v1/data/wildlife-strikes'
curl 'https://himaxym.com/api/v1/data/laser-strikes/state/california'