VECTORDEFENCE DRONE
INTELLIGENCE
INDEPENDENT PERSPECTIVE
THE EVOLVING FIELD GUIDE
Editor

Field guide

How to evaluate defence drone evidence: claims, demonstrations and confirmed use

A practical method for reading sources, checking identities and separating what a demonstration shows from what remains unknown.

In this guide · 4 min read
XQ-58A Valkyrie experimental UAV flying alongside an F-16
U.S. Marine Corps XQ-58A Valkyrie test flight alongside a U.S. Air Force F-16 at Eglin, Florida, 3 October 2023. U.S. Air Force / Master Sgt. Tristan McIntire (opens in a new tab)

Start with the exact claim

A source can be authentic and still support less than a headline suggests. A manufacturer can accurately announce a first demonstration without establishing production readiness. A ministry can accurately announce an order without confirming delivery. The first task is to write down the specific proposition being assessed: which vehicle, which configuration, which event and which date?

Keep a claim small enough to check. The sentence a new autonomous system is operational contains several separate assertions. The vehicle must be identified, the autonomous function described and the meaning of operational established. If only one of those elements is evidenced, publish that element with its limit. A short, precise statement is more useful than a broad claim supported by a narrow source.

Different sources answer different questions

A supplier is well placed to describe its offered product, but has a commercial interest in its presentation. An operator can establish that it received or used equipment, but may disclose little about configuration or performance. An independent assessment can examine a programme critically, although its findings apply to the period and scope studied. Source type is context for interpretation, not a universal score of truthfulness.

GAO's technology-readiness guidance is useful because it asks how maturity has been demonstrated, rather than treating a label as sufficient evidence. It does not certify the systems on this site. We use that distinction as an editorial question: what was actually assessed, under which conditions, and by whom? [1]

What each source can establishSource types are context, not automatic credibility scores.
EvidenceMay establishStill ask
Product catalogueThe supplier’s offered configurationHas it been integrated or accepted?
Test accountA described activity and resultWhich conditions and criteria?
Order noticeA defined procurement eventWhat scope is firm?
Operator statementA described receipt or useWhich variant and period?

Trace repetition back to its origin

Five articles may all repeat one press release. That is five publications, but only one original account of the event. Follow links and distinctive wording to identify the underlying announcement. Preserve the original publication date and the event date where they differ. An old demonstration recirculated this week does not become a new demonstration.

A useful evidence note records the publisher, the source's relationship to the programme, the exact claim it supports and any qualification that changes the meaning. If the original is unavailable, say that the account is secondary. Do not describe a report as independently confirmed merely because several outlets repeat it.

Photographs establish context, not every capability

A photograph can show an object in a particular setting when its provenance and caption are reliable. It does not automatically establish the installed internal equipment, the duration of a test, the number in service or the reliability of an autonomous function. A computer rendering establishes a proposed appearance, not the existence of a working vehicle.

The historical test-flight image accompanying this guide is explicitly captioned. It illustrates why date, subject and context matter. It is not an image of every collaborative aircraft programme and should not be used to imply that unrelated systems have reached the same stage. Image reuse must also respect its source's rights and credit requirements.

Worked example: an impressive demonstration video

Suppose a fictional supplier publishes a short video showing a robotic inspection vehicle moving through a facility. The defensible observation is that the published footage depicts the stated activity, subject to checking its origin. It does not establish how many attempts were needed, whether a person intervened off-camera or whether the equipment can repeat the task under other conditions.

A proportionate article would attribute the demonstration, describe its stated scope and identify what was not disclosed. It would seek a dated test account or customer statement before adding stronger status claims. This is not an accusation that the video is misleading. It is a way to avoid asking one piece of evidence to answer questions it does not address.

Keep confidence attached to individual facts

Do not give an entire vehicle a single confidence label and then assume every specification is equally secure. Identity may be well established while quantity, payload or programme stage remains uncertain. Two sources can disagree because they discuss different variants or dates. Record both qualified statements before deciding whether they truly conflict.

The RAF's June 2025 Protector announcement is a useful historical example: it described service entry while also discussing further work towards full operating capability. Reporting the actual milestone is more accurate than replacing it with an unrestricted claim of completion. [2]

Leave a trail another reader can follow

The final article should let a reader reproduce the reasoning. Link the original source, attribute the claim, identify the configuration and state the unresolved question. When new evidence arrives, update the reference profile and retain dated news as a record of what was known then. A correction should explain the changed fact rather than hide the earlier uncertainty.

For concepts and reported prototypes, publication can still be valuable. Explain who proposed the system, what evidence of hardware exists and which claims remain aspirational. Unknown does not mean false, and an attractive concept does not mean a fielded capability. Both distinctions are necessary for an honest, growing library.

Publication history

  • — Published an illustrated field guide with worked examples and source references.

Suggest a correction to this entry

Follow the evidence

Source types identify who makes the claim. Programme milestones and supplier statements should be read in context.

  1. GAO: Technology Readiness Assessment Guide, 2020 (opens in a new tab)Government / assessment methodology
  2. RAF: Protector enters service, 16 June 2025 (opens in a new tab)Government / operator

Related news

News archive
Airborne drone-carrier proof of concept19 Nov 2026

Schiebel demonstrates S-100 carrying and relaying a smaller drone

Schiebel and OSIRIS Poland say an S-100 helicopter drone released an inert HSCD-1 in forward flight and relayed its control and video link. The Austrian proof of concept did not demonstrate an armed intercept or customer acceptance.

Source: 18 Nov 2026Read
Drone-supply policy target19 Nov 2026

Ukraine sets goal to double drone supplies in 2027

Ukraine's defence minister told EU counterparts that the country plans to double drone supplies in 2027 and seek joint European production. This is a policy target, with no published baseline, signed order or delivery schedule in the announcement.

Source: 17 Nov 2026Read
Air and ground live-fire demonstration17 Nov 2026

US Army demonstrates air and ground drone strike concepts in Germany

At Grafenwoehr, Army developers demonstrated a TRV-150 smoke drop, a R3NEGADE-mounted HIV3 drone launcher and a separate Javelin-equipped ground robot. This was a live-fire technology demonstration, with no reported procurement or service acceptance.

Source: 16 Nov 2026Read
Counter-UAS field experiment16 Nov 2026

US Army reports MANOWAR acoustic counter-drone trials in South Korea

A US Army unit and South Korean partners field-tested a distributed acoustic sensor network for detecting and locating small drones. The Army reports a September classification metric, while scaled production and a formal programme remain goals.

Source: 13 Nov 2026Read
Air and ground logistics experiment14 Nov 2026

EDA and Swiss forces test air and ground drones for urban logistics

At ALPEX LOG, European and Swiss teams assessed UAV resupply and reconnaissance alongside wheeled and tracked ground vehicles for supply and casualty evacuation. EDA reports an operational experiment, not a procurement or accepted configuration.

Source: 13 Nov 2026Read
Counter-UAS operational experiment14 Nov 2026

Lithuanian Army tests counter-UAS systems at Vanguard exercise

Lithuania says Vanguard 2026/II put detection, tracking and neutralisation technologies through practical trials with companies from several countries. The army will use the evidence for further testing and possible pilot purchases.

Source: 12 Nov 2026Read
Uncrewed surface-vessel unit11 Nov 2026

Royal Navy commissions first autonomous surface-vessel squadron

The Royal Navy commissioned 1st Autonomous Squadron in Portsmouth and demonstrated Kraken K3 Scout and Rattler boats under remote control. The new unit is an organisational milestone; the source gives no operational deployment result.

Source: 9 Nov 2026Read
Counter-UAS interceptor assessment10 Nov 2026

US Army assesses FD-1 interceptor and operator training in Poland

An Army air-defence battery assessed the FPV FD-1 interceptor at Ustka and developed training and engagement procedures for massed drone threats. The report describes an internal assessment, not a fielding decision or verified intercept rate.

Source: 9 Nov 2026Read
Maritime autonomy demonstration10 Nov 2026

AUKUS reports cross-domain trials at Maritime Big Play Pacific

Australia says the three AUKUS partners demonstrated aerial, surface and large underwater autonomous systems around San Diego, including ISR, contested resupply and anti-submarine scenarios. The release reports exercise outcomes, not platform-specific acceptance.

Source: 9 Nov 2026Read
Frontline logistics control8 Nov 2026

Ukraine says multi-drone control is in use for frontline logistics

Ukraine's defence ministry says a frontline unit is using one interface to route and supervise groups of cargo UAVs without satellite navigation. The reported doubling of deliveries is a unit claim, with no fleet size or independent comparison published.

Source: 6 Nov 2026Read
Counter-UAS competition7 Nov 2026

US Army opens expeditionary counter-drone fly-away kit competition

An Army competition accepting submissions from 6 November seeks portable systems for detecting and defeating small drones at austere sites. Prizes and a future exercise are planned; no winner, contract or fielded kit has been announced.

Source: 6 Nov 2026Read
Counter-UAS procurement framework6 Nov 2026

NATO opens counter-drone marketplace with 16 selected solutions

NATO says its new counter-UAS marketplace lists 16 competitively selected solutions from five firms under multi-award frameworks. Listing enables faster Allied procurement; it is not an order by every member or acceptance of every configuration.

Source: 5 Nov 2026Read
Counter-UAS procurement5 Nov 2026

Ondas reports Ranger and Iron Drone Raider orders in Europe and Asia

Ondas says it secured counter-drone contracts for portable Ranger systems with an unnamed European police organisation and Iron Drone Raider interceptors with an unnamed Asian government customer. Deliveries and acceptance were not reported.

Source: 4 Nov 2026Read
Airborne anti-submarine trial3 Nov 2026

TEKEVER and General Dynamics report AR5 sonobuoy detection trial

TEKEVER and General Dynamics Mission Systems–Canada say AR5 deployed sonobuoys at REPMUS 2026 and their processing chain detected and tracked underwater contacts. The companies do not give target identities, ranges, trial dates or independent naval acceptance.

Source: 2 Nov 2026Read

Explore the connections