Europe has funded the counter-drone response and left the sensing voluntary

On the evening of 18 September, Luxembourg Airport stopped flying. Unmanned aircraft had been observed moving from the vicinity of the field to over the aerodrome itself, operations were suspended around 21:00, an inbound flight from London went to Liege instead, and traffic resumed in the early hours of Saturday ^[6]. The airport then said the thing that matters more than any of it: it did not have details on where the drones may have come from or who was operating them ^[6].

That sentence is the European counter-drone problem in one line. The response capability is being funded. The sensing that would tell anyone what to respond to, and to whom, is not being required.

What the closures actually establish

Three incidents this year make the pattern legible. Berlin Brandenburg closed both runways and its airspace for roughly 45 minutes on 23 September after a drone was seen over the northern runway, reopening at 21:05 with arrivals diverted to Dresden and Leipzig; no attribution followed ^[7]. Munich suspended flights for about an hour on 30 May after two pilots reported a possible drone, and emergency services found no threat to the public – the sighting was never confirmed ^[8].

Note the detection method in each case. At Berlin the trigger was a person looking up and seeing something ^[7]. At Munich it was two flight crews reporting a possible object ^[8]. Those are legitimate reports and the operational decisions were defensible. They are also not measurements. One publisher put the consequence precisely: whether a series of reports means three drones or three sightings of one drone is exactly the sort of question these incidents almost never answer ^[7].

The contrast case is instructive. Germany did formally attribute an earlier explosive drone attack at Leipzig/Halle Airport on 4 August to Russia ^[7]. Attribution became possible when there was physical evidence to work from. Where there is only a sighting, there is nothing to attribute, and an airport pays the operational cost anyway.

The plan names detection and requires none of it

The Commission’s Action Plan on Drone and Counter Drone Security, COM(2026) 81 final of 11 February 2026, is not negligent about sensing. It is built on three pillars – Prepare, Detect, Response – and its detection section describes a multi-sensor approach integrated through AI-powered command-and-control software, naming passive and active radar systems, software-defined radars, acoustic sensors, LiDAR and thermal infrared cameras ^[4]. It proposes using 5G networks and Integrated Sensing and Communication to detect non-cooperative drones that carry no identification, and references CIMACT as a single air display system allowing detection and identification of potential threats in near real time ^[4].

What it does not do is require any of that of anybody. A search of the document for mandated detection requirements or performance standards for airports or critical infrastructure returned none; what it offers instead is voluntary guidance and stress-testing initiatives for willing Member States ^[4]. The timelines are development timelines, not compliance deadlines: a cellular-detection testing initiative in Q2 2026, an EU counter-drone centre of excellence in Q1 2027 ^[4]. A recommendation on counter-drone performance requirements is due in Q4 2026 ^[3]. A recommendation is not a standard.

Set that beside the one measure described as mandatory. The Drone Security Package due in Q3 2026 is to ensure mandatory registration for all operators of drones above 100 grams ^[3]. That is a reasonable rule and it binds precisely the population that was never the problem: operators willing to register. The non-cooperative drone over a runway is unaffected by a registration duty, which is why the plan separately proposes technical means to distinguish authorised from non-cooperative activity ^[3]. The duty is enforceable; the sensing that would make it meaningful is encouraged.

Where the money went

The funding split is the clearest statement of priority. Of the EUR 400 million announced in February, EUR 150 million is for aerial surveillance and drones used by Member States and Frontex in joint operations, and EUR 250 million supports Member States purchasing counter-drone systems directly, as Commissioner Magnus Brunner set out ^[5]. The larger share buys effectors. Reporting at the time also noted that the previously trailed drone wall concept was absent from the package, reframed as a drone initiative ^[1].

The wall itself has not gone away. The European Drone Defence Initiative is described as a multi-layered system with interoperable counter-drone capabilities for detection, tracking and neutralisation, with initial operational capability expected by the end of 2026 and full functionality by the end of 2027 – while open questions remain concerning sensors, effectors, networking and command and control ^[9]. The same analysis makes the economic argument that drives all of this: deploying fighter jets and expensive air-to-air missiles to counter drones is operationally possible but fiscally unsustainable ^[9].

On 28 September the European Council approved five European Defence Projects of Common Interest, among them DECODER, a drone and counter-drone programme, making them eligible for funding through the European Defence Industry Programme ^[2]. The framing reported was building and scaling drones and the technology to stop enemy drones ^[2]. Research programme history points the same way and also shows the correction under way: an ENACT review in June found EU-funded counter-UAS work moving from isolated detection or mitigation components towards integrated end-to-end capabilities covering detection, tracking, identification, situational awareness and mitigation ^[10].

What to do

• Separate the two questions before buying anything. Detection answers whether something is there; identification answers what and whose. A system that does the first and not the second will still close your airfield ^[4].
• Treat a visual report as a cue, not a measurement, and design the corroboration path in advance – radar track, RF detection, acoustic, or camera – so that the decision to suspend operations has something to resolve against ^[7][8].
• Record what the sensors saw, per sensor and unmerged, including the negative result. Munich’s value as a data point rests on emergency services finding no threat, which is only useful because it was recorded ^[8].
• For infrastructure operators outside Europe, read the Q4 2026 recommendation on counter-drone performance requirements when it lands and treat it as the de facto specification, since it is the nearest thing to one that will exist ^[3].
• Do not expect registration rules to reduce incursions. Mandatory registration above 100 grams governs cooperative operators; the detection problem is defined by the ones who do not identify themselves ^[3].

The R3KONX view

Counter-drone work is usually discussed as an effector problem – what you shoot, jam or net. The harder half is electromagnetic and it is upstream: establishing that an object is present, classifying it, and attributing it well enough to justify a response. That is a sensing and signal-processing discipline, and it is the half Europe has chosen to encourage rather than require.

For this region the lesson transfers without the politics. Malaysian airports, ports and substations face the same asymmetry: an incursion costs an operator a closure whether or not anything was really there, and the capability that would settle it is a procurement nobody is obliged to make. The question worth asking before the next suspension is not what we would do about a drone. It is what we could prove about one, and from which sensor.