A reusable, fully-autonomous multi-mission combat drone. It sees the fight with a full EO/IR + designator, SAR & SIGINT sensor suite, navigates like a pilot — no GPS, no datalink — survives the laser + HPM + kinetic counter-UAS triad, and slots into the Air Force network only when it chooses. Built by the thousand on autonomous automaker lines. And it comes home — to refuel, rearm, re-fly. Only the munition is spent.
A fully parametric OpenSCAD model — monocoque shell, pusher, V-tail, retractable tricycle gear, faired & guarded EO/IR turret — rendered to a dimensioned plan & profile, four orthographic/iso views, a 13-part exploded assembly, a structural half-section cutaway, and a 360° assembly animation — all generated deterministically from one source file. Concept GA; dimensions are targets [TBR]. ↓ GA source (.scad) · ↓ Assembly source (.scad)




The same parametric model, driven by an explode factor — every component separated along its own assembly axis. Internal subsystems are concept massing [TBR].
Each photoreal frame is generated from the CAD render itself (image-to-image) — the model keeps the exact geometry and re-lights it as real hardware. Concept visualization beside the technical drawings above.



WILDFIRE isn't inspired by the interview — it's a literal engineering readout of it. Twenty design drivers, each anchored to a verbatim quote with a timestamp.
RF links and fiber "are probably going to go away." WILDFIRE flies the whole mission with the radios off — autonomy does everything.
"Not off GPS or any radio… you look out the window and drive to the place you need to go." Vision-only nav + terminal target ID.
"Made in a Ford or GM or John Deere or Caterpillar factory… a thousand a day, not a thousand a decade." Steel, rivets, single-stage presses.
"Almost impossible to build something that can stop all three of those at the same time." WILDFIRE is built to be that impossible problem.
"A $300 chip is not the thing driving that price." Best-in-class edge AI, treated as a rounding error against the airframe.
The airframe, seeker, and compute are recovered every sortie. Reusability is the program's #1 KPP — and its economic engine.
A one-way airframe is the "total folly" Palmer warns against. WILDFIRE flies the entire loop 50+ times — the only thing it ever leaves behind is the munition. Here's the full breakdown.
Palmer's insight: any single counter-drone effector is beatable cheaply — but forcing a drone to beat all three at once is "really, really hard." WILDFIRE imposes exactly that problem on the enemy's C-UAS designer.
Low-cost ablative / reflective skin + thermal mass + optional body spin spreads and sheds directed-energy dwell.
Faraday-enclosed avionics, an optical internal data bus, and transient protection on every aperture.
A lateral solid-propellant divert motor kicks WILDFIRE clear of the interceptor's lethal radius in the final instant.

Iran's drone kills came from exactly what WILDFIRE is built to beat: a GPS-spoofed RQ-170, a SAM-killed Global Hawk, and $30M Reapers lost one at a time. WILDFIRE flies vision-only / GPS-denied (nothing to spoof or jam), ingresses low and low-signature under the radar horizon, and saturates batteries that engage only 4–6 targets at once — at a ~31:1 cost-exchange ($4M interceptor vs a $130k reusable airframe that flies home).
WILDFIRE's autonomy core is the same vision-navigation stack flown in the AI Grand Prix (presented by Anduril): forward camera in, control out — no GPS, no LiDAR, no human in the loop.
The full ISR · targeting · EW · threat-warning suite, all fitted simultaneously — best-in-class, US-origin or close-allied, MOSA-modular. This is the no-compromise build. The cost-disciplined fielded baseline tiers down to a mission-kit subset (the 4-sensor designator turret alone is a ~$225k quote); full SWaP-C, alternates, and ITAR notes live in the engineering dossier.
A 2-axis stabilized multi-sensor turret under the nose — daylight EO, MWIR thermal, SWIR, laser designator/rangefinder and a spot tracker in one ball. On an expendable drone this dies on every shot; on WILDFIRE it is recessed, ring-guarded, and recovered every sortie.


Trillium HD59-MLVS — MWIR + LWIR + EO + SWIR with a STANAG laser designator + rangefinder. Recessed, faired & ring-guarded so the PTZ turret survives belly / Skyhook recovery — the seeker comes home and is reused.
Vantor Raptor vision/terrain-referenced fix + ANELLO X3 photonic-gyro IMU + M-code/CRPA + low-SWaP star tracker. Navigates and targets with no GPS, no radio.
IMSAR NSP-3 — all-weather, day/night synthetic-aperture imaging plus ground moving-target indication through cloud, dust, and obscurant.
ASI SNITCH — passive HF–18 GHz signals intercept with single-platform 3D direction-finding and emitter geolocation — no GPS, no datalink.
Laser-warning receiver + IR missile/hostile-fire approach warning — detects the laser or interceptor and cues the terminal divert (DDR-11) in the last instant.
Multiple wide-FOV global-shutter cameras feeding the Thor-class brain — the "pilot's eyes" for navigation, autonomous Skyhook recovery, and terminal target ID.
Palmer's rule is "buildable in a car factory" — by the people who already work there. WILDFIRE's DFM-simple airframe — stamped steel, rivets, <120 parts — drops onto the existing lines of domestic automakers (Ford · GM · Stellantis), where skilled crews and robots build it side by side: automation takes the dull, dirty and dangerous; people do the skilled assembly, operation and judgment. Retool, don't build greenfield. Domestic first; the allied auto base is the fallback. American jobs, at rate.
Modern auto body shops already pair skilled crews with welding & stamping robots. WILDFIRE's <120-part, generous-tolerance design retools straight onto them — robots take the dull and dangerous; people do the skilled work.
Every airframe is born from a model-based digital thread; in-line machine-vision augments human inspectors — flagging defects fast while people make the call. Full unit-level traceability.
Surge across domestic automakers first; the allied auto base — the "consolation prize" Palmer names (Japanese automotive workers, DDR-20) — is the fallback. Reuse multiplies effective fleet capacity on top.
The airframe shell isn't stamped — it's a glass/basalt composite monocoque, wet-compression-molded on matched steel tools in ~120–180 s, with steel frames over-molded at the hardpoints — "soft shell, hard bones." It's the same robot + skilled-crew press cell modern automakers already run, so it drops onto the existing line and scales to ≥1,000/day on ~3–9 presses (~12–24 steel tool sets).
Retool an existing automaker body-shop line into an 8-station flow — <120 parts, generous tolerances, model-based digital thread. One line clears ~100+ aircraft/day; surge to ≥1,000/day across ~9 lines and ~12–24 steel tool sets. No greenfield arsenal.
| Phase | Line configuration | Build rate |
|---|---|---|
| Prototype · Y1 | 1 pilot cell, hand-assisted | 1–2 / month |
| LRIP · Y2 | 1 retooled body-shop line + 1 press cell | ~10–30 / month |
| Rate-1 · Y3 | 1 full line + ~3 press cells | ~150–300 / month |
| Rate-2 · Y4 | 2–3 lines + ~6 press cells | ~500–900 / month |
| Full rate · Y5 | multi-line, ~9 lines / ~12–24 tool sets | ≥1,000 / day surge |
Concept point design, Rev B.1 — closed at 175 kg MTOW. Figures are engineering targets/estimates; the full traceability, budgets, and risk register live in the dossier.

The blue-side answer Palmer calls "the top of my pinnacle": a fast, recoverable kinetic interceptor that goes out, kills, and comes home to be refueled, rearmed, and reused. Same autonomy core, same manufacturing doctrine, same open network.
Fury is a teammate, not a rival — WILDFIRE's real competitor is the expendable affordable-mass class: Anduril's Barracuda (and ALTIUS loitering munitions). Barracuda's bet is that "affordable mass" means a ~$150–200k airframe you throw away every shot — airframe, seeker and compute gone per engagement. WILDFIRE keeps the same automaker-line economics but recovers all of it — only the munition is spent — collapsing cost-per-effect from ~$150k/shot to ~$3.2k/sortie. Full comparison + honest caveats in the dossier →
| Dimension | WILDFIRE (reusable) | Barracuda-class (expendable) |
|---|---|---|
| Spend model | Reusable — airframe + seeker + compute recovered; only the munition spent | Expendable — whole vehicle destroyed per shot |
| Cost per effect | ~$3,210/sortie amortized + fuel + munition | = full unit cost every shot (~$150–200k/effect) |
| Unit flyaway | ~$130k | Family target <~$200k; FRP goal ~$150k |
| Role breadth | Multi-role — 25 kg swappable bay, re-role between sorties | Single-mission per round (sensor vs kinetic split across rounds) |
| Range | 1,500–2,500 km radius / 12–20 h — and it returns | Barracuda-250 >370 km; -500 >930 km — one-way |
| Autonomy / nav | Vision-only, GPS-denied, datalink-optional | Lattice autonomy; GNSS-class nav, datalink common |
| Survivability | Hardened + terminal divert, then recovers | None required — meant to die on target |
| Best at | Persistent, repeated, multi-role, recoverable-sensor missions | One-way deep strike & hyperscale saturation |
Barracuda spends its whole value every engagement. WILDFIRE re-spends only the munition — its decisive number is ~$3.2k/sortie, not its $130k flyaway. Even a few reuses crush the exchange ratio.
Expensive EO/IR+designator, SAR, SIGINT and Thor-class compute are recovered every sortie — the expendable class re-buys equivalent sensing and processing every single shot.
Both fly the same EW fight, but a jammed expendable that misses is simply lost — a jammed WILDFIRE terminal-diverts and recovers its seeker and compute. No operator-link single point of failure.
A 25 kg swappable bay re-roles a recovered airframe across ISR · SAR · SIGINT · EW · strike. Matching that breadth with one-shot rounds means buying — and expending — many more units.
1,500–2,500 km radius and 12–20 h on station — that returns — vs a single outbound profile. Persistent ISR + strike + EW, not one effect then gone.
Honest: one-way deep penetration of peer IADS, zero recovery infrastructure, full outbound endurance, hyperscale saturation (~300k planned), and raw simplicity/$. Reuse wins for repeated, survivable-airspace missions — not universally.
Anduril's Fury (YFQ-44A) is the exquisite high-end of the USAF CCA spectrum — an uncrewed, autonomous ~$25–30M, Mach-0.95, AMRAAM-armed jet (first flight Oct 2025, in production at Arsenal-1). Both WILDFIRE and Fury fly themselves — no pilot aboard either; the human commands from the rear. WILDFIRE is the low-end affordable mass beneath it — ISR, EW, decoy, designation and strike-support at ~200× lower unit cost. They don't compete; they team through a common, government-owned autonomy ecosystem — affordable mass under the exquisite. Full comparison + upgrade path in the dossier →
| Dimension | WILDFIRE (AGP-1) | Fury (YFQ-44A) |
|---|---|---|
| Layer | Low-end affordable mass | High-end exquisite CCA (Increment 1) |
| Role | Reusable multirole: ISR · strike-support · EW · decoy · designation | Fighter-class UCAV: air-to-air, manned-unmanned teaming |
| Propulsion | ~35 hp heavy-fuel pusher (turboprop dash opt.) | Jet turbofan (~Williams FJ44-class) |
| Speed | 100–130 kt cruise | ~Mach 0.95 |
| Range / endurance | 1,500–2,500 km radius · 12–20 h | Not publicly disclosed |
| Autonomy | Vision-only, GPS-denied, datalink-optional (flies + recovers radios-off) | Semi-autonomous, on-the-loop; ArsenalOS / Lattice |
| Payload / weapons | 25 kg modular bay; A/A optional (U4) | 2× AIM-120 AMRAAM |
| Cost | ~$130k flyaway · ~$3.1k/sortie | ~$25–30M |
| Basing | Runway-independent: rocket-sled + Skyhook | Runway / short-field |
Collaborative multi-ship behaviors + optional cockpit/operator tasking (quarterback model). Software-only and RF-optional — degrades to autonomous single-ship when jammed.
Adopt the government autonomy reference architecture so competed mission-autonomy (Lattice/Hivemind) loads onto WILDFIRE and tactics ship as software — no airframe recert. Highest-leverage, lowest-cost import.
Electronic-attack + deception in the open bay — make a Group-3 drone look like a larger platform. A cheap survivability multiplier, recovered with the airframe.
Optional lightweight air-to-air pylon (short-range AAM) so WILDFIRE isn't pure prey — modular, human-on-the-loop. Baseline stays unarmed and cheap.
A separate faster jet/turboprop sibling (like BACKFIRE) for the contested ISR-dash sliver — reuses the autonomy core, manufacturing & network. Never the baseline.
Ruggedized-laptop task/upload C2 and a few lightly-trained maintainers, no fixed base — reinforces the ≤30-min turnaround. The link is convenience, never required for recovery.
This package was run through an adversarial engineering review. Here's exactly what closes and what's still an open, managed risk — because that's what a real program looks like.
Reusable. Vision-guided. Survivable. Mass-producible. Every requirement traces to a line you actually said. The full engineering dossier — requirements matrix, subsystem designs, budgets, risk register — and the business & DoD-contract plan are one click away.