The aircraft gets the photograph. The research list gets the headache.

On 5 October 2026, the US Federal Aviation Administration announced $12 million for 19 listed drone research and testing projects. The eye-catching entries include long-range cargo aircraft and large uncrewed systems. The more revealing entries are less cinematic: a common altitude reference, secure identity checks, passive radar, surveillance for long routes, mission-coordination software and links between drone traffic services and conventional air traffic systems.

Taken together, the list makes a blunt point. Scaling drones is not mainly a matter of building more aircraft. It is a matter of making many aircraft, operators and services agree on who is flying, where they are, what they are allowed to do and how they avoid everything else.

What the FAA actually funded

The FAA announcement says the awards were made through its Unmanned Aircraft Systems Broad Agency Announcement programme, with companies working alongside FAA-designated UAS Test Sites. The 19 project descriptions include ground and onboard detect-and-avoid systems, drone conspicuity, passive radar, satellite communications, common-altitude tools, identity and approval verification, urban airspace reservations, integration between drone and traditional traffic-management systems, and software for coordinating missions and resources.

Those are project descriptions, not published results. The announcement does not provide performance data, completion dates or evidence that every technology is ready for routine use. It confirms that the work has been selected and funded for research and testing.

The programme itself is built around demonstration and validation. The FAA’s BAA programme page cites the FAA Reauthorization Act of 2024, which made $6 million available in each fiscal year from 2025 through 2028 for matching funds and test-range work on technologies the agency considers important to safe integration. In other words, the test is the point. An award is the beginning of an evidence process, not the final stamp at the end.

The list reads like a safety stack

Safety stack is our editorial shorthand, not an FAA term. It describes several layers that have to work together before dense or long-range drone operations can become ordinary.

Layer one: see other aircraft and stay apart

Several awards concern detect-and-avoid, visibility or surveillance. General Atomics is listed for ground systems to help drones avoid other aircraft. OneRadio is testing passive radar. Skydweller is working on onboard avoidance for large drones. Lighthouse Avionics is developing surveillance tools for long-range routes, while two other projects address drone visibility.

This cluster matters most when the remote pilot cannot personally scan the entire operating area. But “detect” is not the same as “avoid”, and neither word describes a universal system. A useful safety case has to explain which cooperative and non-cooperative aircraft can be sensed, at what range, with what confidence, how tracks are maintained, what manoeuvre follows and what happens when the sensor or link degrades.

The FAA’s current detect-and-avoid documentation tool shows how specific that review can become. It asks applicants to describe aircraft size, speed, altitude, operating time, right-of-way procedures, emergency procedures and avoidance manoeuvres, among other factors. The FAA also says operations are reviewed case by case and may require an operational evaluation. A successful sensor test therefore cannot be stretched into blanket approval for every aircraft or route.

Layer two: agree on the same airspace picture

Another group of projects is about shared information rather than onboard hardware. ANRA Technologies appears in the list for a common altitude reference, links between drone traffic systems and traditional airport traffic systems, and connections among air-taxi operations, vertiports and flight procedures. Aerial Vantage is listed for managing reserved urban airspace.

The common-altitude item sounds almost comically basic. It is not. Two systems can each report a precise height while referring to different surfaces, datums or conventions. Precision without a shared reference can produce confidently incompatible answers.

The broader coordination model is UAS Traffic Management, or UTM. The FAA describes UTM as a collaborative low-altitude ecosystem built from regulatory requirements, technical capabilities and interoperable services. Its functions can include flight planning, authorisation, surveillance and conflict management. The agency’s vision relies primarily on automated data exchange through application programming interfaces, rather than a controller speaking individually to every drone pilot.

That is the scale problem in one sentence: voice does not multiply easily; interoperable data might. But APIs only help if the participating systems share definitions, update quickly, remain available and handle disagreement safely.

Layer three: prove identity without confusing it with permission

DroneUp is listed for a secure way to verify drone identity and approval to fly. ANRA has a separate project using AI tools to identify different types of drones. These projects sit beside, but should not be confused with, the existing US Remote ID rule.

The FAA defines Remote ID as the ability of a drone in flight to broadcast identification and location information that other parties can receive. That helps establish an identity and location signal. It does not, by itself, prove that a particular flight has every permission it needs, that its route is conflict-free, or that the person receiving the broadcast can trust every other system in the chain.

Identity, authentication, authorisation and surveillance answer different questions. Who are you? Is the credential genuine? Are you allowed to perform this operation? Where are you now? A scalable system needs those answers to connect without collapsing them into one reassuring green icon.

Layer four: keep the operation connected and coordinated

Censys Technologies is listed for testing satellite communications for long-range drone flights. ResilienX is working on software to coordinate missions and resources. Devorto’s project concerns long-range cargo service. These entries point toward the operational layer: communication paths, service availability, scheduling, contingencies and the handover from one system to another.

A long-range aircraft can have an excellent airframe and still face a weak operation if its command link, navigation input, surveillance service or coordination platform becomes unavailable. The useful question is not whether one component works in a demonstration. It is what the whole operation does when one component does not.

What the $12 million does not prove

The award announcement is evidence of research priorities and funded work. It is not evidence that the projects have completed testing, met a published performance threshold, received operational approval or entered commercial service.

That distinction is especially important because company and project names can make research sound like deployment. The FAA’s own programme language is careful: demonstrate, test, evaluate and validate. Until results and approval records exist, claims about reliability, coverage, superiority or readiness should remain open.

There is also a jurisdiction limit. These awards concern integration into the United States National Airspace System. They do not set UK or EU operating rules, and they do not establish what a British pilot or operator may do.

Five questions for the next “drone scale” announcement

1. How will the aircraft detect and avoid other traffic? Ask which traffic can be sensed, which cannot, and what happens after detection.
2. Do every participant’s position and altitude mean the same thing? A shared reference is infrastructure, not a formatting detail.
3. How are identity and authority verified? A broadcast identifier, an authenticated credential and permission for an operation are separate layers.
4. What happens when communications or a service fails? Look for degraded modes, contingency procedures and evidence from realistic tests.
5. What result has actually been published? Separate funding, demonstration, validation, regulatory acceptance and routine service.

The aircraft is only the visible part

The FAA’s list is interesting because it spends so much attention on things a spectator will never see. A shared altitude reference will not make a dramatic launch video. Neither will an API exchange, an authentication check or a software service that quietly prevents two operations from colliding.

That is probably a sign of maturity. Aviation becomes routine when the invisible parts are dependable enough to be boring.

For now, these 19 projects are a map of questions the US system wants answered. The next useful evidence will be the test methods, limitations and results—not another promise that scale is just one more aircraft away.

Scope note: This article analyses an FAA funding announcement and current FAA programme material, rechecked on 13 October 2026. It does not report independent flight tests, predict project outcomes or state that any funded technology has operational approval.