Each case distinguishes event, deployment and publication dates. Research prototypes and manufacturer reports do not establish general commercial performance.
1987 / JP
R-50: commercialising agricultural unmanned aviation
CONTEXT
Yamaha encountered weight, cost and handling problems with a counter-rotating design and shifted to a single main rotor.
REPORTED OUTCOME
In 1987, the R-50 demonstrated spraying with a 15 kg payload; a monitored run of 20 aircraft went on sale that December.
LESSON FOR THE NEXT FIELD
Simplifying the mechanism and monitoring early users supported introduction.
LIMITATIONS
Manufacturer history; not evidence of present-day multirotor performance or approvals.
2018 / CN
Multi-aircraft control is not automatically swarm intelligence
CONTEXT
DJI documented multi-aircraft control for the MG-1P, with a November 2018 user manual.
REPORTED OUTCOME
The manufacturer specifies simultaneous control of up to five aircraft: a documented product capability rather than only a research concept.
LESSON FOR THE NEXT FIELD
Separate concurrent operation from cooperative control and emergent swarm intelligence.
LIMITATIONS
Product documentation does not establish field productivity, safety outcomes or permission for concurrent operation.
2019–2021 / FR
Vineyard drift depends on the nozzle
CONTEXT
PULVEDRONE compared drone and ground-equipment drift within an experimental steep-vineyard framework.
REPORTED OUTCOME
The poster reports large nozzle-dependent differences; the tested drift-reducing drone configuration had a drift index comparable to a crawler sprayer.
LESSON FOR THE NEXT FIELD
Evaluate aircraft, nozzle and operating conditions together.
LIMITATIONS
These configuration-specific results do not prove biological efficacy, universal equivalence or current permission.
2023 / CA
Turn imagery into reseeding evidence
CONTEXT
University of Saskatchewan research addressed laborious manual counting of early canola seedlings.
REPORTED OUTCOME
A 2023 university account reports a model returning seedling counts and ground-cover estimates to farmers, with further image training underway.
LESSON FOR THE NEXT FIELD
Deliver a quantity that informs a decision, not just attractive imagery.
LIMITATIONS
Research reporting does not independently establish reseeding outcomes, yield gains or transfer to other crops.
2024 / JP
Why a farm tested a ground alternative after drone seeding
CONTEXT
Nagano’s Goto Farm reported labor-saving and yield benefits from drone direct-seeding on 0.4 hectares the previous year.
REPORTED OUTCOME
In 2024, it tested a transplanter-mounted direct-seeding attachment on three hectares, valuing simultaneous fertilizer and herbicide application.
LESSON FOR THE NEXT FIELD
Choose tools by the complete workflow, including combined operations.
LIMITATIONS
One farm’s configuration, not a universal drone limitation; emergence, water management and birds remained concerns.
2025 / AR
Establish vetch before maize harvest
CONTEXT
An April 2025 demonstration at Marcos Juárez broadcast vetch into standing mature maize.
REPORTED OUTCOME
The report records 7.29 hectares, refilling and battery changes, then emergence and cover. Actual application was 31.57 kg/ha against a 25 kg/ha target.
LESSON FOR THE NEXT FIELD
Assess timing gains alongside actual distribution and establishment.
LIMITATIONS
Commercial demonstration, not a randomized ground-equipment comparison; some dates and elapsed-day labels conflict.
2025 / US
Refills and heat constrain spray capacity
CONTEXT
A 2025 Mississippi State Extension report describes field evaluations presented by Auburn University and other researchers.
REPORTED OUTCOME
The Auburn researcher reported refilling and battery swaps about every five minutes in their work, with substantial summer labor demands.
LESSON FOR THE NEXT FIELD
Measure the ground cycle and workload alongside flight time.
LIMITATIONS
A reported experience, not a universal service interval; advertised endurance needs configuration-specific validation.
2025 / ZA
Connect cane measurements, spraying and harvest
CONTEXT
A study published in 2025 linked refractometer readings and the PurEst app to ripening decisions on fragmented smallholder fields.
REPORTED OUTCOME
Sixteen demonstrations used drones and farmer-readable comparisons. Estimated recoverable sugar increased, while cane mass decreased in most trials.
LESSON FOR THE NEXT FIELD
Distinguish crop mass, saleable value and understandable advice.
LIMITATIONS
Contractor availability, fragmentation and harvest coordination limit wider adoption. The year denotes publication, not every trial date.
2020–2025 / TW
Validate two-way rural logistics in stages
CONTEXT
Taiwan’s transport institute began island and mountain logistics validation in 2020, including goods and agricultural produce in both directions.
REPORTED OUTCOME
Technical validation in 2020–2022 progressed to extended service trials in 2023–2025 and business-model validation from 2025.
LESSON FOR THE NEXT FIELD
Treat technical feasibility, sustained service and commercial viability as separate milestones.
LIMITATIONS
The government progress report does not establish produce volumes, quality retention or profitability.
2026 / ID
Indoor swarms: distinguish research from deployment
CONTEXT
In March 2026, IPB University described indoor-farming swarm research intended to back up fixed sensors.
REPORTED OUTCOME
Researchers describe central analysis of imagery and environmental data, inter-drone communications and position coordination.
LESSON FOR THE NEXT FIELD
Specify both the coordination mechanism and maturity level when comparing this with product multi-aircraft control.
LIMITATIONS
A research account, not proof of commercial outdoor spraying swarms, certification or improved profitability.