Artificially generatedProject Stachelschwein: five million roofs against ten thousand drones
Stachelschwein DDS is the open drone shield for Germany. First we listen. Then we catch. Then we take the trash out of the sky.
Project Stachelschwein, internally Stachelschwein DDS, is the decentralised drone shield for Germany. Five million sensors and interceptor drones on suitable roofs, tied together by an open protocol. A wave of ten thousand drones is not answered by one expensive missile. Mass answers it, the kind you can roll out in weeks instead of parliamentary terms.
Every suitable house gets an ear. Some later get a net.
Stachelschwein DDS ties cheap sensors and cheap interceptor drones into air cover that reaches roofs, not just airports. Every suitable house gets a sensor on the roof. Every suitable house can get a defence drone with explosive-free capture gear, nets for example. All nodes run on one shared platform.
Five million sensors and interceptors protecting German airspace together.
On top of that, Stachelschwein adds the next layer: the fire brigade and the federal agency for technical relief get longer-reaching sensors and stronger interceptors. The system is modular from the first day and built so it can later plug into the command and information structures of the Bundeswehr and NATO.
We are not the manufacturer, at least not so far. We start the thing and carry the idea. Anyone who wants to build is welcome. Fast coordination by state agencies is expressly wanted.
Bottom-up. Fast. Cheap. Area cover, not one airport.
Attack waves of ten thousand drones are no longer theory. No single expensive missile answers that. Mass answers it, the kind you can roll out in weeks instead of parliamentary terms. The platform is open to plug-and-play across sensor and drone systems. Only two things are developed centrally and jointly: the standardised protocol and the software that fuses sensor data and hands out interception tasks.
A listener on a roof. Nothing has to fly.
The Stachelschwein Scanner is week one. Interceptors are week later. Missile bags are a research page. You put a box on the roof, or on the balcony, and you look at the sky.
Direct Remote ID has been mandatory in the EU since 1 January 2024 for class-marked C1 to C3 drones in the Open category and for every drone in Specific. A phone app hears that for zero euros. An ESP32-S3 hears it for about 12 EUR. Four MEMS microphones hear the propeller even when the aircraft is dark: fibre-optic FPV, no radio, no RID.
| Node | Price | What it hears |
|---|---|---|
| Phone listener | 0 EUR | Remote ID, app on the phone you already have |
| RID box | ~25 EUR | ESP32-S3, cable, case |
| Acoustic box | ~50 EUR | 4 MEMS mics, ESP32, printed shell |
| Node with Pi 5 | ~150 EUR | Brain plus one ear |
| Five million roof nodes at 150 EUR | 750m EUR | Listening layer, before interceptors, labour, layer two |
An RTL-SDR does not see 2.4 GHz. The R820T2 dies around 1.8 GHz. Infineon XENSIV BGT60TR13C is 15 metres, presence in a room, not city radar. SPEXER and Aaronia are layer two: fire brigade, THW, the troops. The roof gets ears first.
First named product: Stachelschwein DDS Scanner v0.1. Bill of materials, 3D unit, monitor.
Four messages carry the whole sequence. Arming is not one of them.
Only the standardised part is developed centrally. Bring a camera, a microphone array, a radar or an interceptor drone, write an adapter against those messages, and you are part of the network.
| Message | Direction | Payload |
|---|---|---|
| DETECTION | node → cell | Bearing, time, quality. Never a finished position. |
| TRACK | cell → node | Several detections become one track. |
| TASKING | cell → interceptor | Launch clearance, target, role. Never arming. |
| OUTCOME | interceptor → cell | Caught, missed, aborted. |
The Frankfurt simulation already runs this chain: at least three bearings make a track, ten interceptors per inbound, a net, an acknowledgement, no explosion. Those are simulation values, not vendor claims.
Floor four is Müllabfuhr im Himmel, a rubbish truck in the sky. Make the enemy's weapons useless. Later. Nets around drones first. Missiles in a sack is a research page, not the start.
We are not waiting for a decision.
We are citizens who consider a resilient air defence infrastructure necessary, so we are starting to build it. Defence only. No selling of weapon systems. Within the law.
- We install from today.
- We arm nothing before it is agreed with the authorities.
- We create readiness. In an emergency the button belongs to whoever is authorised to press it.
That way the coverage exists before the rules are finished. The responsibilities can grow alongside it. System design from 7 September 2026. First installations: passive nodes, from today. Flight tests and arming only with regulatory clearance. Five million systems as fast as production carries.
The surface is the workshop. The Wunderglobe magazine is the argument. GitHub is the mirror.
The project surface lives on Wunderglobe because the globe already holds Frankfurt as a model and the simulation runs there. The argument that started this sits in the Wunderglobe magazine. The software and protocol part is public under Apache 2.0. Anyone may read it.
Scanner, system, parts, supply chain, home test, cost.
ArgumentThe piece that made the shield public.
SimulationFrankfurt, PLAY GAMES. Or Stachelschwein and Chill.
CodeProtocol sketch, hardware minimums, system design.
The sky over a German roof is a listening room from today. The box costs as much as a decent dinner. The rest is mass.
| Claim | Basket | Source |
|---|---|---|
| Direct Remote ID required in the EU since 1 January 2024 for C1, C2, C3 (Open) and every drone in Specific. Standard EN 4709-002. | Fact | EASA, EUR-Lex 2019/945 and 2019/947 |
| C0 drones under 250 g often need no Direct Remote ID in Open. A RID scanner then sees empty sky. | Fact | EASA Open category table, 1 January 2024 |
| Raspberry Pi 5 is built in Pencoed, Wales. Design: Cambridge. | Fact | raspberrypi.com, sonypencoed.co.uk |
| System design from 7 September 2026. Public repo Apache 2.0. | Fact | Wunderglobe magazine, GitHub LICENSE |
| Three bearings for a track, ten interceptors per inbound. | Fact as simulation value | Wunderglobe, src/stachelschweinMath.js |
| Five million systems, ten thousand drones as the opposing picture. | Target / thesis | Wunderglobe magazine, 7 September 2026. Not a production pledge. |
| Roof node RID plus acoustics 30 to 50 EUR without a Pi, with Pi 5 closer to 150 EUR. Five million at 150 EUR = 750 million EUR listening layer. | Hypothesis, medium confidence | Street prices on /cost, not a quote, no VAT file |
| Müllabfuhr im Himmel, nets around missiles: research and a joke. Not the start. | Hypothesis / stance | John, 9 September 2026 |
- Project surface Stachelschwein DDSScanner, system, parts, cost, home test. English default at /stachelschwein.https://wunderglobe.com/stachelschwein
- Wunderglobe magazine: Project StachelschweinThe argument from 7 September 2026. Five million, ten thousand drones, stance.https://wunderglobe.com/mag/en/projekt-stachelschwein
- GitHub: stachelschwein-drone-defense-systemProtocol, hardware minimums, system design. Apache 2.0.https://github.com/JohnMcGrinsey/stachelschwein-drone-defense-system
- EUR-Lex 2019/945Delegated regulation on drones, the basis for Remote ID in the EU.https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX:32019R0945
- EUR-Lex 2019/947Implementing regulation on the operation of unmanned aircraft.https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX:32019R0947
- EASA: civil dronesOpen category, C classes, Direct Remote Identification.https://www.easa.europa.eu/en/domains/civil-drones
- Frankfurt simulationwunderglobe.com/play, PLAY GAMES, Sim Stachelschwein.https://wunderglobe.com/play
Cut: 9 September 2026. McGrinsey writing skill. John's lines on mass, the missing decision and Müllabfuhr im Himmel left standing.
