An automated defence line thirty metres out in grassland — further in brush and in forest, because the terrain and the fuel set the real distance, not a price list. A continuous wet band that forms itself the moment a fire front approaches: on its own water, its own solar power, with a fluorine-free agent, and with nobody present.
The engineering is done. The costs are modelled. What we are looking for is the partner who builds the first one.
That is the whole idea. Everything below this page is the detail an engineer, a fire officer or an insurer will ask for — but you do not need any of it to decide whether the idea is worth building.
Every wildfire protection product on the market answers those two questions — and almost all of them answer both the same way: at the house, and once the fire is already there. Plot the field on those two axes and an empty square appears.
| When it acts | Who | The limit |
|---|---|---|
| Months ahead | Matrix Wildfire · statutory firebreaks | The treatment degrades, and nothing can be done in the moment itself |
| Minutes ahead | Frontline · waveGUARD · per-house systems | Only the surface of the house — by then the fire is on the plot |
| Afterwards | Team Wildfire · insurer response fleets · aircraft | Someone has to arrive, with people and machines, at the worst possible moment |
| In the moment | The Standoff Line | Nobody covers this today |
The same field again, but asked differently: does it start on its own, does it need somebody or something to arrive, and what does it actually keep the fire off. One row has a tick in every column that matters, and that row does not exist as a product yet.
| System | Starts | Needs a person to work | Keeps the fire off | |||
|---|---|---|---|---|---|---|
| BY ITSELF | BY HAND | THE APPROACH | THE PLOT | THE BUILDING | ||
| Statutory firebreak, cleared or ploughed strip | · | · | · | ✓ | · | · |
| Retardant laid down months ahead — Matrix Wildfire, CitroTech | · | · | · | ✓ | · | · |
| Cannon towers around a development — Medi XXI WUIProtect, Cantudo | ✓ | ✓ | · | ✓ | · | · |
| Exterior sprinklers on one house — Frontline, waveGUARD, House Water Dome | ✓ | ✓ | · | · | ✓ | ✓ |
| Ember mesh and home hardening — Chapter 7A, EmberPro, TECT | · | · | · | · | · | ✓ |
| Suppression after ignition, by vehicle — Team Wildfire, the fire service | · | ✓ | ✓ | ✓ | ✓ | ✓ |
| Aerial attack | · | ✓ | ✓ | ✓ | · | · |
| Research lines — STME FIRE (PAF), ADAI Coimbra at Travessas | ✓ | ✓ | · | ✓ | · | · |
| The Standoff Line | ✓ | ✓ | · | ✓ | ✓ | ✓ |
The question is what each measure needs when the fire arrives, not how it was built: every one of them is created and maintained by people, so that is not counted. A dot in both start columns means the measure is passive — a cleared strip or a screened vent is simply there and nobody operates it. The approach is the ground 30–100 m out, the plot is the ten metres around the house, the building is the envelope itself. The research lines answer the same question we do, but you cannot buy one.
And where they are ahead of us, which is most of it. Medi XXI has a granted patent, twenty years of installations and a delivered community scheme of forty towers. Frontline reports sixty-one houses and ninety-six per cent survival in the 2025 Los Angeles fires. Matrix treats twenty-four kilometres in one pass and works at night and in wind. Every one of them has something we do not: a built reference. What the table shows is not that we are better today — it is that the column combination in the last row is still empty.
We did not invent this idea, and we say so first. In Spain, Medi XXI has built cannon-based perimeter defences around developments since 2006 and holds a granted patent. In France, STME FIRE runs a research programme on exactly this. In Portugal, university researchers ringed a village with twelve sprinkler columns in 2020. In the USA, two funded companies build barriers at industrial scale.
Not one part of this system is new. What is new is that nobody has put them together. We are the first to combine what already exists into a single system with three layers of defence — around a settlement, or at smaller scale around one property: the outer line that holds the front at a distance; the plot and the garden, defended against embers that travel for kilometres, using the irrigation system that is already there or is built along with it; and finally the buildings themselves. One controller, one tank, one set of pumps. Small heads instead of a few big cannons, dual-triggered, energy-independent, and an agent that already meets the 2030 European ban on fluorinated foams.
A brass impact head with a 15 m radius every 20 m, so the circles overlap into a 22 m unbroken wet band. At 10 m radius the band would have holes — and a fire goes through the holes.
A satellite fire alert arms the line at roughly 10 km. A local UV/IR flame detector fires it at roughly 300 m. Fast and precise, without choosing between them.
Own tank, own pump, solar with battery and an auto-start generator. At least four hours of operation. Mains water pressure and grid power are the first things a wildfire takes away.
A Class A foam measured below 1 mg/kg fluorine, GreenScreen Silver, and already on the US Forest Service qualified products list. It meets the 2030 EU PFAS ban today.
The flame detectors say which arc the front is coming from, so only that sector runs, plus two poles of margin. The water lasts two to three times longer.
Start time, pressure, water used, temperature, wind, runtime. Insurers price what they can model — so the line produces the evidence from its first day.
The line stands dry. A weekly self-test cycles the valves, the solar array keeps the battery full, the tank stays full. Nothing runs, nothing is wasted, and the same pipework can irrigate.
A satellite fire alert wakes the system. The pump pressurises, the sectors report ready, and the owner gets a notification — wherever they are.
Solar-powered UV/IR flame detector nodes along the line identify not only that a front is coming, but from which arc.
Only the threatened sector starts, plus two poles of margin. The overlapping heads build a continuous 22 m band; foam is dosed at 0.1–1%. In strong wind the controller switches to a coarser droplet so less of it drifts away.
At least four hours from its own tank. The active sector follows the front as it passes, and the band keeps cooling the ground after the flame has gone — which is what stops a re-ignition.
An event report: start time, runtime, pressures, water and agent used, temperature and wind. This is the file an insurer can model — and the reason the evidence exists from the first fire onwards.
| Above ground | Below ground | The plant room |
|---|---|---|
| 3–5 m galvanised steel pole on a concrete footing Brass impact head, 15 m radius, 25–30 L/min at ~4 bar Solar UV/IR flame-detector nodes Inner ring: roof and eave sprinklers — the ember layer |
HDPE main, buried 0.45–0.5 m Metal pipe wherever it surfaces Sectioning valves per arc A collection channel that returns unevaporated water to the tank |
Tank, sized to the vegetation type Electric pump plus diesel auto-start backup Solar array, LiFePO4 battery in a fire-rated cabinet Foam proportioner 0.1–1% and concentrate tank Controller: 4G, satellite alert receiver, weather and level sensors |
Standard fire-brigade couplings on the tank, so an engine can draw from it — required by name in the Valencian building code, and the fastest way to make the local fire service an ally rather than an obstacle.
Whatever the source, the line always runs from its own tank — the source only refills it, because a wildfire takes the mains away first. Which source it is depends on the geography of the site and on what the local rules allow, and usually it is more than one.
Roof and hardstanding collection into the same tank. Cheap and unregulated almost everywhere — but in an arid climate the roof area will not fill a tank sized for four hours, so there it is a supplement, not the supply.
Where a treatment works is within reach, this is the cheapest large volume there is, and it is already being done: the community scheme in the Turia park near Valencia runs on recycled water. Needs a permit and a discharge agreement.
Independent of the grid, which is the point, but it needs its own pump, its own power and a measured yield — a borehole that is fine for irrigation may not refill a fire tank between two runs. Abstraction is licensed almost everywhere.
The simplest connection and often the only one available. It fills the tank between events and never feeds the line directly: mains pressure and grid power are the first two things to go when a fire arrives.
Works anywhere and needs no source of its own. The tank carries standard fire-brigade couplings, so a water tanker, an agricultural bowser or a fire engine can fill it — the fallback where there is no rain, no treatment works and no borehole, and the way the tank is topped up again after a run. It fills the tank; it never feeds the line, because that would put a driver back into a system built to work without one.
And the water that does not evaporate is collected and used again. A shallow channel inside the line returns what runs off — still carrying agent — to the tank, and the system puts it back through. Not all of it comes back: wind drift and evaporation take a share, and on a slope some of it leaves the plot. But the part that returns is the difference between one run and three.
| Per pole, per fire event | Per house, full protective ring | Community line, 1 km |
|---|---|---|
| ~2.9 m³ water 11–19 L of concentrate ~0.9 kWh 440–600 m² protected |
€31–82k grassland (13 poles) €37–98k brush (19 poles) €62–163k forest (44 poles) 7,700 / 15,200 / 25,200 m² protected |
$121–251k = €1,500–7,400 per home The version a street can actually afford |
Figures from our own hydraulic, energy and cost model, at European price levels. Distances: 30 m for grassland, 50 m for brush, a double row at 50 + 70 m for forest; roughly double on a slope the fire climbs.
And what it is not. Up to 90% of homes lost in a wildfire are ignited by wind-driven embers, not by the flame front — and embers fly over any band. The line stops the front and takes away the radiant heat; ember-resistant vents, screens and a wetted roof stop the rest. We supply both, and we will never claim that one replaces the other.
There is no photograph of this system anywhere, because nobody is selling it yet — which is the whole reason this page exists. So here it is drawn: what one pole is made of, and what the ring does when a front arrives from one side.
No installed reference exists — ours or anyone's. We have not built one yet, and we say so throughout this page. But the reason there is no photograph to show is not that the idea failed: it is that this square of the field is still empty. That is what we are asking a partner to help us change.
Wind-driven embers fly over any wet band and land behind it: on the lawn, in the shrubs, in the gutter, on the roof. Up to 90% of the homes lost in a wildfire are ignited this way, not by the flame front. A line that holds the front and leaves the plot behind it dry is a line that fails. So the design has three layers, not one — and the second and third cost money too, which we would rather price in now than discover on site.
Holds the flame front and takes away the radiant heat before it reaches the plot. Thirty metres out in grassland, fifty in brush, a double row at fifty and seventy in forest, and roughly double on a slope the fire climbs — the terrain and the vegetation decide, and we survey before we quote. This is the part that does not exist on the market today, and the part this page is mostly about.
The ember landing zone. An inner ring of low heads wets the ground and the planting inside the line, off the same tank and the same controller, before the front arrives. The first metre and a half from the wall stays non-combustible: no bark mulch, no shrubs against the cladding, no firewood stacked under the eaves.
Every opening screened with 1.6 mm non-combustible mesh — attic and gable vents, soffits and eaves, crawlspace vents, dryer and extractor outlets, the roof-to-wall junction. Class A roof covering, closed eaves, gutters kept clear, and roof and eave sprinklers on the same controller as the line.
The mesh figure is not ours to invent: 1.6 mm (1/16 in) non-combustible mesh on every opening is what California's Chapter 7A requires for new building in wildfire zones, and what the IBHS ember research supports. We build to the published standard rather than to a number of our own.
And the ground between the ring and the line? It is not wetted, and we will not pretend otherwise: watering ninety metres of plot would need a tank nobody is going to buy. That ground is kept lean instead — mown, no ladder fuels, nothing stacked against a tree. It is also why we quote the shortest standoff the fuel allows: 30 m in grass, 50 m in brush, a double row at 50 and 70 m in forest, and roughly double on a slope the fire climbs. Every extra metre of standoff is another metre of plot that somebody has to keep in that condition.
This makes the system more expensive, and we say so before you ask. The line on its own is the cheapest part of a defence that actually works. The inner ring, the mesh and the roof sprinklers add to it — screening a house typically runs from a few hundred to a few thousand euros depending on how many openings it has, and the inner ring is priced on the plot, the planting and the roof rather than from a table. What we will not do is quote a line that leaves the ground behind it burning.
We can design it, size it, write its control software and manage its construction. What we cannot do is finance the first installation. That is the whole proposition, stated plainly.
Fire-protection and irrigation contractors · manufacturers looking for an application · utilities and railways · municipalities and fire districts · insurers and MGAs who already pay for mitigation · resorts, estates and industrial sites in fire-prone regions.
A 100–200 metre demonstration line, instrumented, filmed, measured. If there is interest, we can install it anywhere, at the site the interested party designates. That is the cheapest credible proof in existence, and it is what turns a drawing into evidence.
And the honest list of what we do not have. No installed reference. No stamped fire-engineering design. No licence in any jurisdiction. What we do have is a concept worked through to cost level, the competitor and regulatory picture in writing, the software that makes it a defence line rather than a sprinkler, and a project director who has delivered before. We would rather open with that list than be found out holding it back.
An economist by training who has taken projects from the first meeting to handover. Former general director (Tổng giám đốc) of an export poultry plant in Thanh Hóa province, Vietnam — built with Hungarian technology and Hungarian capital on a six-hectare site, and inaugurated on 24 October 2019 in the presence of the Hungarian and Vietnamese agriculture ministers and the Hungarian ambassador. Decades of international project management across two countries and two languages, and for many years a regular interpreter for high-level delegations of both governments, including the Vietnamese National Assembly delegation to Hungary in 2022.
He is not an engineer and does not present himself as one. What he brings is the thing most often missing on a build: that the project actually starts, that the parties pull in one direction, that the money and the schedule hold, and that somebody answers for the handover.
Software and systems engineer: the sensing, the trigger logic, the sector control, the telemetry and the event log — and the application that lets an evacuating owner see that the line is alive. A former flight instructor, which is where the habit comes from of designing for the moment when something fails rather than the moment when everything works.
In this system the software is not an accessory. The difference between a sprinkler and a defence line is what decides when it runs, how much water it spends, and what it can prove afterwards.
If you build fire protection, irrigation or infrastructure, and you have a perimeter that needs defending — that is the conversation we want. We answer in every language this page is written in.
The full concept study, the cost model and the competitor analysis are available on request, in any of the sixteen languages on this page.