Paper 1
No Trend, Only Season
A decade of FAA UAS sighting reports shows no significant trend; the signal is weekend-driven and decoupled from public attention.
Finding No trend, 2015–2025 (p = 0.814); +62% weekend effect (t = 20.1).
UUAS Research Series
The U.S. drone airspace, measured from the record up.
Original research on FAA drone-sighting and registration data — a rebuilt, openly-licensed dataset and four studies of where drones are flown, where they are reported, and why the two diverge.
21,765
records rebuilt
2014–26
full record
4
papers (of 7)
CC BY 4.0
open license
Papers
Paper 2
Exposure Without Encounter
The first state-level cross of sighting reports against remote-pilot density: where drones are owned is not where they are reported.
Finding Operator density does not predict sightings once population is controlled (51 jurisdictions).
Paper 3
A Complete, Reproducible Record (2014–2026)
A cleaned, gap-corrected reconstruction of the FAA UAS Sighting Reports, recovered from delisted quarters, with the full pipeline.
Finding 21,765 records; recovers the delisted 2014–2019 era; corrects a published error.
Paper 4
Counting What Cannot Be Counted
The federal instruments that count U.S. drones and pilots were built for accountability, not measurement — and cannot see the active base.
Finding Effective fleet ≈ a third of the registration tally; the active workforce is unknowable from the record.
Forthcoming: Paper 5 (registry fleet composition), Paper 6 (waiver & COA analytics), Paper 7 (approach-path structure) — pending an FAA FOIA production and further analysis.
Data & code
The reconstructed dataset — 21,765 FAA UAS sighting reports (2014–2026), recovered from delisted quarters, with the full build pipeline and a provenance manifest — is released under CC BY 4.0 with a permanent DOI. Everything is reproducible from three commands.
Zenodo deposit → doi.org/10.5281/zenodo.21444168
Zenodo deposit → doi.org/10.5281/zenodo.21444168