Everything else
fails to off.
This fails to
deployed.
A reflective shroud that drops over a home's ignition points when fire arrives — and drops anyway when the power dies and the water stops. No pump. No grid. No person on site.
Live stateStowed · armed
The defense that disappears
when you need it
Homes in the wildland–urban interface rarely fall to a wall of flame. They fall to wind-borne embers that travel ahead of the front and slip in through eaves, soffits, vents, and window gaps — and to radiant heat that ignites surfaces and breaks glass.
The systems sold to stop this — exterior sprinklers, foam, retardant — all share one dependency: water and power. In a major event, both go first. Hydrants drop the mains; utilities cut the grid to reduce ignition risk; pumps stop.
This isn't a worst case. Pressure collapse is documented in the Tubbs, Camp, and Marshall fires. Homeowners watched rooftop sprinklers run for minutes, then trickle, then stop. And every one of those systems fails the same direction: to off — to unprotected.
Lose water pressure or grid power and the defense simply stops. Tanks and wells only partly cover the gap, and the failure leaves the structure exposed.
The barrier is held back by a restraint that releases under heat. Lose power, water, comms, or a sensor and the restraint still lets go — so the shroud drops.
Four ideas, one shroud
A flexible reflective barrier sits stowed in a cassette at each vulnerable point. A spring is pre-loaded toward down. A heat-sensitive restraint is the only thing holding it up.
Fail-safe to deployed
Deployment follows from losing the restraint — not from gaining power. So loss of grid, water, comms, or sensing doesn't disable the shroud. It deploys it.
Stored energy, not drawn energy
The work of dropping the barrier is loaded into a spring at install. Nothing has to be powered or pumped at the moment of fire.
Many signatures, one trigger
Convective heat, incident radiant flux, or a direct ember — any one releases the restraint. Radiant response can drop the shroud before the flame front even arrives.
Distributed, not central
One assembly per region, each triggered locally. No shared bus, no shared conduit. If one fails to deploy, the others are unaffected.
Every way in gets its own guard
The structure is protected region by region. Pick a point of entry to see what guards it.
Eave & soffit
The most common ember path. Wind-borne embers lodge in the open underside and ignite framing from above. A shroud unfurls across the eave line, closing the gap and reflecting radiant load off the fascia.
The trigger is an OR gate
The restraint is built from independent release mechanisms wired in mechanical OR. The first fire signature to appear lets go. Loss of integrity from any cause frees the spring.
One event, start to shielded
No controller runs this. It is a chain of physical events — each one the cause of the next.
Radiant flux rises
The front is still away, but its radiant heat reaches the structure first and begins loading the restraint.
First signature wins
The radiant element parts — or an ember strikes, or convective heat melts the link. Whichever comes first.
The restraint lets go
Loss of restraint integrity frees the spring. No signal is sent; nothing is switched on.
Stored energy drives down
The pre-loaded spring unwinds the barrier along its guides — mechanical, not metered.
Heat turned away
The reflective face throws radiant load back; the covered region is closed to embers.
Tension keeps the seal
A weighted leading edge holds the barrier taut against wind through the event.
Three layers do three jobs
Reflective outer layer · 222
An aluminized or metallized face that throws incident radiant heat back before it can raise the surface to ignition.
Refractory core · 224
A fiberglass or ceramic-fiber substrate that insulates and holds dimensional stability at high temperature.
Backing layer · 226
Protects and stabilizes the laminate so it survives stowed for years and unfurls clean.
The material is a disclosed option — the invention is the deployment, never the cloth.
Against the field
Each existing family does part of the job. Only one does it autonomously, region-wide, and fails toward protected.
| Stegophylax | Ember vents | Retardant coatings | Active sprinklers | Manual wrap | |
|---|---|---|---|---|---|
| Deploys without grid power | Yes | Yes | Yes | No | Yes |
| Deploys without water | Yes | Yes | Yes | No | Yes |
| Autonomous — no person on site | Yes | Yes | Pre-applied | If powered | No |
| Covers a whole region | Yes | Aperture only | Surface | Wets area | Yes |
| Multi-signature trigger | Yes | Heat only | — | Sensors | — |
| Fail direction | Deployed | Sealed (vent) | Washes off | Off | Human |
Assessments reflect typical configurations of each family; particular products vary.
The filing
The separation
The four adjacent families fall away on a single positive limitation and one functional distinction:
- Ember vents seal an aperture; the shroud covers a region.
- Coatings and sprinklers depend on chemistry, water, and power — severed by “independent of external power and fluid.”
- Aluminized wrap is the material, deployed by hand. The invention is the passive deployment, never the cloth.
The honest catch
Interior fire curtains already drop on a trigger. So novelty can't rest on “deploys when hot.” It rests on the combination: exterior ember context, a reflective barrier, a fully passive multi-signature trigger with no power path, and a fail-safe, distributed architecture — and that is the first target for the professional search.
The hard questions
Depends on the trigger. A eutectic link is consumed and swapped during service, like a sprinkler fusible link. A shape-memory variant resets after it cools. The barrier and cassette are serviceable either way — the assembly is built to be re-armed, not discarded.
The barrier rides captive guides and is pulled by a weighted leading edge, so deployment is positively driven rather than wind-dependent, and the refractory textile holds tension once down. High-wind behavior is exactly what bench and field validation must characterize before any claim of performance.
The logic is deliberately deploy-biased: a false deploy is the safe-side error. Thresholds are set per region to the expected exposure, and re-arming is a low-cost service action — by design the system would rather drop early than miss.
No. In the armed state the barrier is stowed in its cassette and the opening is clear. It only covers the region on a fire trigger, so day-to-day light, view, and egress are unaffected.
The optional supervisory layer that reports state does use power — and it sits entirely outside the actuation path. Deployment never waits on it. Cut its power and every assembly still deploys on its own.
Not yet. The provisional is drafted at full scope but not filed, so there is no priority date and nothing to license today. This page is staged behind a pre-filing gate; see Status below.