Industries / Food, farming and forestry / Agriculture and farming
Crop dusting
Crop dusting, also called aerial application, is the work of spraying crop chemicals and spreading fertilizer and seed over fields from planes, helicopters and drones.
1921
On 3 August 1921 Army pilot Lieutenant John A. Macready flew a modified Curtiss JN-4 "Jenny" biplane over a six-acre (about 2.4 ha) grove of catalpa trees near Troy, Ohio, and dusted it with lead arsenate to kill caterpillars; the FAA's historian calls it the first use of the airplane for crop dusting.
35
Milestones achieved across 6 domains.
8
To watch for: announced, not yet confirmed outside.
12
Open deliverables. Nobody has shipped them.
In one minute
Where it started
In 1921 an Army pilot dusted an Ohio grove with insecticide from a converted training biplane, and by 1925 Huff-Daland had built a plane for the job and started a dusting business in the US South, the forerunner of Delta Air Lines. After World War II, surplus planes and pilots spread the work: the US counted over 5,000 aircraft in farm use by the early 1950s, and New Zealand began spreading fertilizer on hill country by air in 1949. Purpose-built crop planes followed from 1950, designed so the pilot could survive a crash.
Where it stands
By the trade group NAAA's 2019 survey, about 1,560 US aerial application businesses and about 3,400 pilots treat 127 million acres (about 51 million ha) a year, about 28% of US cropland counting repeat treatments; Brazil had 2,866 farm aircraft at the end of 2025. In the US 83% of fixed-wing crop planes had turbine engines by 2019, while 69% of Brazil's were still piston-powered; pilots now fly satellite-guided passes, and spray drones are spreading under permits granted case by case, with a little over 13,000 declared to Brazil's aviation agency for farm work by October 2026. The European Union has banned spraying from the air since 2009 except by derogation, and in December 2025 its Commission proposed easing the ban for listed types of drone.
What is open
Much of the record of this work is scattered or locked in each maker's format: where the spray went, how many aircraft fly, what each drone permit allows, and which towers and wires pilots hit. The open jobs here turn that into public data, maps and models: a checkable open drift model, one table of measured drift trials, a count of crop planes from aircraft registers, a converter for spray logs, and maps of unlisted towers and of vineyards too steep for tractors. None of them requires anyone to fly or spray.
Domains
AchievedWatch forOpen
Domain
Progress
Last achieved
Watch for
Open
Aircraft
This work is designing and building the planes and helicopters that carry seed, fertilizer and crop chemicals low over fields, the gear that releases the load, and the features that let pilots survive the job.
6 · 1 · 2
Last achievedA crop plane on ethanol2004
Watch forNozzle-by-nozzle control in flight
2
openSpray drones
This work sprays or spreads fields from aircraft with no pilot on board, from battery multirotors to converted helicopters and fixed-wing planes.
6 · 1 · 2
Last achievedA US sprayer start-up folds2025
Watch forAn uncrewed R44 sprayer
2
openApplication and drift
This work decides how seed, fertilizer or pesticide leaves the aircraft, how evenly it lands across each swath (the strip one pass covers), and how much of it drifts beyond the field.
6 · 1 · 2
Last achievedEPA updates its aerial model2025
Watch forDrone drift in the middle
2
openMapping and guidance
This work keeps each pass of the aircraft the right distance from the last, turns the spray on and off at field edges, and records where everything went.
5 · 1 · 2
Last achievedSpray rates follow a map2009
Watch forLive wind for the pilot
2
openPilot licensing and rules
This work decides who may fly and spray from the air, in which aircraft, over which land and with which products, and how each job is recorded.
5 · 2 · 2
Last achievedEU drone relief withdrawn2024
Watch forDrones exempted from EU ban
2
openSafety
This work keeps pilots, loaders and people near treated fields from harm: crashes into wires and towers, poisoning by the chemicals carried, and spray that lands where it should not.
7 · 2 · 2
Last achievedPilots knew the wire2023
Watch forFewest crashes on record
2
open
Advances to build with
1 research note
Open in crop dusting
3 of 12 open deliverables
Open · Aircraft
Counting the world's crop planes
A yearly count of agricultural aircraft by country, model, engine type (piston or turbine) and year built, made from public civil aircraft registers rather than surveys, published with the code to rerun it.
Built with what exists
The FAA's downloadable aircraft register; Brazil's ANAC aircraft register (RAB), which Sindag uses; other national registers where published; Python and pandas; a list of agricultural models (Air Tractor, Thrush, Ag-Cat, Ipanema and others) to catch aircraft not flagged by category; a web-browsing AI agent to find and fetch each country's register file (see the Note "The jobs agents repeat").
Open · Spray drones
Where drones meet crop dusters
A public estimate of how often a spray drone working one field shares the air with a crewed crop-dusting plane working another nearby. Collect a season of ADS-B tracks (Automatic Dependent Surveillance–Broadcast, the position signal many aircraft send) from aircraft of agricultural types, mark their low passes over fields, then run simulated drone jobs on neighboring fields and count how often the two come within set distances.
Built with what exists
The FAA's public aircraft registry, to list every aircraft of an agricultural make and model with its ADS-B address; the OpenSky Network's historical ADS-B database, free for research; the open-source traffic Python library to clean tracks; USDA's Cropland Data Layer to mark fields; BlueSky, an open air-traffic simulator with conflict detection, for the simulated drone jobs.
Open · Application and drift
Finish the open AGDISP
A Python version of AGDISP, the Forest Service's aerial drift model, that anyone can script and run in batches, with a public test suite showing it gives the same downwind deposit as the official program.
Built with what exists
The AGDISP 8.29 source in that repository; Python and NumPy; AI coding agents to translate the code and write the tests, as in AWS's cloud emulator that agents wrote and tests alone checked (see the Note "Code no person read"); the official Windows AGDISP to make the reference outputs.
Where it sits
Broader
