From simulation to armored flight
Six phases. Each one has an exit test that a stranger could verify, a budget estimate, and a kill criterion. Nobody flies free until the tethered hours are in. Nobody wears armor until free flight is boring. Today the program is at the start of Phase 0.
Simulation and flight-control stack
Physics model of the 5-engine body with engine lag, prone and upright aerodynamics, pilot G model. Controller inner loop, thrust allocation and envelope protection running on the real flight computers in hardware-in-the-loop. 500 injected faults, 0 crashes in sim.
Exit: HIL rig recovers every single-engine-out from hover and cruise. Pilot standard and medical drafted with a flight surgeon.
Tethered hover rig
Exoframe, fuel system, 5 engines, gimbals, arc core, ground station, gantry. Pilot in a plain flight suit, helmet, airbags. First engine runs on the stand, then on the frame, then with a pilot under the gantry. 50 tethered hours.
Exit: 10 consecutive 3-minute tethered hovers with the controller holding position within 30 cm, including 5 engine-out cuts performed by the ground crew without warning.
Free hover over foam
Untethered, under 3 m, over the foam pit, inside the geofence, pilot-incapacitation protocol armed. Then run-on transitions to 10 m/s along a 50 m strip.
Exit: 20 flights, zero controller abort surprises, auto-land demonstrated from 3 m with hands off the throttles.
Low cruise and the dead-zone transit
Climb-outs through the 3–120 m band with forward speed, chute armed, prone transitions above 60 km/h, 150 m circuits. Pilot centrifuge-qualified before this phase. Operate inside FAA Part 103-style limits where possible: single occupant, daylight, under 102 km/h, over unpopulated ground.
Exit: 25 circuits, one deliberate chute deployment from 150 m on a ballast dummy, G-LOC ladder validated in the centrifuge against the real pilot.
Armored MK-I
Composite metal foam torso and helmet through a research partnership, 48-zone shell sensing, full HUD, 250 km/h envelope. Mass goes up ~5 kg; every Phase 1–3 test repeats.
Exit: armored sortie flying the mission profile end to end, with a 5.5 g prone evasive pull, pilot fully conscious, controller never past “assisting”.
Autonomy research
Only after Phase 4: can the controller fly the suit empty for ferry and test, and can anything learned from logged flights safely tune the supervisory layer? This is where HQ, Hylaq’s local-first on-device intelligence research, would be evaluated. It is a research question, not a plan with a date.
Kill criteria · the program stops and is redesigned if
- Any uncommanded engine shutdown that the controller did not predict from engine-health data, twice.
- Any G-LOC or greyout event in flight. Ever. One is a redesign of the ladder.
- Any fire outside a test stand.
- Pilot core temperature over 39 °C or CO over 50 ppm in a nominal flight.
- HIL and real flight disagree on an engine-out recovery by more than 20% in any axis.
- Chute test fails to fully inflate within 2 s.
Regulatory path
United States: FAA Part 103 covers a powered ultralight under 254 lb (115 kg) empty weight, 5 US gal fuel, under 55 kt (102 km/h), single occupant, daytime, not over congested areas, with no licence or airworthiness certificate. The MK-I dry weight (~80 kg without pilot and fuel) and 20 kg fuel (~6.6 US gal, slightly over) put it near the line; Phases 1–3 are designed to fit with a 19 L fuel load. Beyond 102 km/h or with armor, it becomes an experimental aircraft and needs an FAA airworthiness certificate and a pilot.
Canada: Transport Canada basic ultralight category and a pilot permit. Everywhere: Phases 0–2 are ground and tethered operations, which need a landowner and insurance, not an aviation authority. Get an aviation lawyer involved before Phase 3.
About $350k to an armored first flight
| Line | Estimate | Notes |
|---|---|---|
| 5 micro-turbines + 1 spare | $70–85k | P550-class engines run $12–14k each with ECU and starter-generator. |
| Exoframe, gimbals, fuel system, shell | $35k | Two frame iterations. Carbon work outsourced, machining in-house. |
| Avionics, compute, sensors, telemetry | $20k | Automotive SoCs, industrial IMUs, LIDAR puck, RTK GNSS. |
| Helmet, HUD, life support, G-suit, chute, airbags | $30k | Waveguide HUD module is the expensive item. |
| Test infrastructure: gantry, stand, pad, foam, ground station | $45k | One-time. Reused for every future suit. |
| Fuel, consumables, spares, insurance, centrifuge time, flight surgeon | $40k | Over 18 months. |
| CMF partnership and panels (Phase 4) | $100k+ | Research licence and qualification. Could be sponsored. |
| Total to armored first flight | ~$350k | Excludes salaries. Two full-time engineers for 30 months is the real cost. |
The team we need
- Propulsion engineer who has integrated micro gas turbines (RC jet or UAV background is perfect).
- Controls engineer who has shipped a real flight controller (PX4/ArduPilot contributors, eVTOL alumni).
- Mechanical/composites for the exoframe and shell.
- Flight surgeon or aerospace physiologist, part-time, owns the pilot standard and the G-LOC ladder.
- Test director who has run a hazardous test program and will say no to the founder.
- Pilot: fit, calm, centrifuge-qualified, and not the person who signs the cheques.
This month
- Publish the program with honest status labels. Done.
- Flight surgeon conversation: validate the ladder and the pilot standard.
- Quote 6 engines. Lead times are 3–6 months; order gates everything.
- Stand up the physics sim and the flight-control repository.
- Find the pad and the landowner.
- Open the funding conversation with the budget above and the Phase 1 exit test as the milestone.