Here's the complete analysis as a standalone HTML page you can drop straight into your website. All key engineering estimates are flagged where they're based on assumptions that need tank/field validation.
```html
Solar Seastead USV — 1:4 Scale Model Engineering Study
1:4-Scale Solar Seastead USV — Engineering Study
Froude scaling, weight & power budgets, speed estimates, thruster redundancy,
foiling analysis, stabilization hardware, salt-spray protection, rescue concepts,
competitor landscape and cost model.
Answers at a glance:
• Scale factor λ = 4 → lengths ÷4, areas ÷16, weights ÷64, speeds ÷2
• Target model weight: 36,000 ÷ 64 = 562.5 lb (budget below lands at ≈555 lb — feasible but tight)
• Batteries (30%) = 169 lb ≈ 7.5–8 kWh LiFePO₄ (≈6 kWh usable above 20% floor)
• Solar ≈ 800–850 W on the 11-ft triangle (≈950–1,000 W if grown to 12 ft)
• Hotel load ≈ 55–60 W; motors ≈ 100–150 W night / 250–350 W day
• Cruise speeds ≈ 2.7–3.0 kt night, 3.5–3.9 kt day (≈70–90 nm/day)
• Foiling: possible as a sprint (~5 kt, ~5–6 h, ~30 nm) — not an efficiency win
• 6-thruster redundancy: statistical loss-of-steerage probability per 2-week mission < 0.01%
1. Froude Scaling (λ = 4)
| Quantity | Scaling law | Full scale → Model |
|---|
| Length | ÷ λ | ÷ 4 |
| Area (solar, foil, wetted) | ÷ λ² | ÷ 16 |
| Volume / displacement / weight | ÷ λ³ | ÷ 64 |
| Speed | ÷ √λ | ÷ 2 |
| Time (wave period, roll period) | ÷ √λ | ÷ 2 |
| Force | ÷ λ² | ÷ 16 |
| Power | ÷ λ3.5 | ÷ 128 |
| Wave height / sea state | ÷ λ | ÷ 4 |
2. Model Dimensions (feet & inches)
| Feature | Full scale | 1:4 model |
|---|
| Triangle side | 44.00 ft | 11 ft 0 in |
| Leg length (vertical fin) | 21.5 ft | 5 ft 4½ in |
| Foil chord | 8.5 ft | 2 ft 1½ in |
| Max thickness (NACA 0035) | 2.975 ft | 8 15/16 in |
| Trailing-edge cut | 0.5 ft | 1½ in |
| Design draft (submerged half of 14.5 ft) | 7.25 ft | 1 ft 9¾ in |
| Freeboard, deck underside to WL | 7.25 ft | 1 ft 9¾ in |
| Waterplane area, all 3 legs (chord × thickness) | ≈76 ft² | ≈4.75 ft² |
| Buoyancy at "desired" waterline (27,500 ÷ 64) | 27,500 lb | ≈430 lb |
| Camera mast (your chosen non-scale value) | — | 4 ft |
Note: At the 562.5 lb target you are above the scaled 27,500-lb waterline rating
(430 lb), so the model floats ≈5 in deeper than the "50% immersed" ideal — exactly as the
full-scale boat would at 36,000 lb. Either accept the deeper draft or trim weight toward
~430–480 lb. Everything below assumes 562.5 lb as instructed.
Your instinct about the model being a brutal stress-test is correct: a 1:4 model cruising
in real Caribbean seas experiences dynamically similar conditions to the full-scale
seastead in seas 4× larger. Control algorithms proven on the model are being validated far
outside the parent vessel's normal envelope.
3. Weight Budget (target ≤ 562.5 lb)
| Item | Est. weight (lb) | Notes |
|---|
| 3 legs (AL 5083, ~1/16″ shell, frames, watertight bulkheads) | 120 | wetted area ≈100 ft² total; 1/16″ plate = 0.89 lb/ft² |
| Triangle frame (2″×2″×⅛″ 6061-T6 angle + purlins) | 75 | ≈60–70 ft of extrusion |
| Heave plates (3 bolt-on) | 25 | |
| Active stabilizers (2 foils + tails + actuators) | 40 | see §9; trim to 25 lb if needed |
| 6 × thrusters + ESCs | 30 | ~5 lb each installed |
| Batteries (30% target) | 169 | LiFePO₄ in leg voids; fits comfortably |
| Solar array (~850 W lightweight) + cabling | 35 | flexible/semi-rigid, §6 |
| Electronics: Starlink Mini (2.4), Pi stack (2), cameras + mast (8), AIS (1.5), nav LEDs (1), converters/fusing (4) | 19 | |
| Wiring, connectors, glands | 12 | |
| Rescue gear (rope, floats, funnel) | 6 | |
| Fasteners, sealant, anodes, paint, misc | 25 | |
| TOTAL | ≈556 | Margin ≈ 6 lb (1%) — very tight |
Recommendation: shave to ~540 lb for margin: thinner frame web, lighter stabilizers
(−15 lb), batteries at 27% (−17 lb). Every pound matters at this margin.
4. Battery Sizing
- 30% of 562.5 lb = 168.75 lb ≈ 169 lb (76.5 kg)
- LiFePO₄ pack-level energy density ~95–110 Wh/kg → 7.3–8.4 kWh; call it ~8 kWh installed
- Usable above 20% floor (your rule): ≈6.0–6.4 kWh
- Leg cross-section is ≈148 in², so prismatic cells + BMS fit easily in the lower third of each leg (≈56 lb/leg).
- Design note: sealed legs need pressure relief / gas venting for the BMS, and per-leg temperature sensing.
Battery-only endurance (no sun, 60 W hotel): ~5.4 kWh to propulsion → at 150 W that's
~36 h × 3 kt ≈ ~105 nm of dark autonomy.
5. Power Budget: Hotel Load & Motor Allocation
| Load | Average W | Notes |
|---|
| Starlink Mini | 30 | ranges 20–40 W; biggest consumer |
| Raspberry Pi / CM4 + Hailo accelerator | 8–12 | vision processing running |
| Cameras (2–3 streams) | 4 | |
| AIS Class-B transponder | 3 | bursty transmit |
| LED navigation lights (night) | 4 | |
| DC-DC conversion & misc losses | 6 | |
| Total hotel load | ≈55 day / ≈60 night | |
Solar harvest
- Triangle 11 ft side → gross area 52.4 ft²; ~88% packing → ≈46 ft² = 4.3 m²
- Rigid slim mono panels (~200 W/m², IEC 61701 salt-mist rated): ≈850 W
- All-flexible ETFE/SunPower option: ≈165 W/m² → ≈700 W (lighter, pricier, less durable)
- If enlarged to 12 ft side: 62.4 ft² gross → ≈5.1 m² packed → ≈950–1,000 W.
Worth doing if you can find ~10 lb of weight budget.
- Caribbean yield: 5.5–6.0 peak-sun-hours × 0.75 derate (heat, spray film, MPPT) →
3.4–4.5 kWh/day. Flat-horizontal mounting is nearly optimal at 18°N latitude.
Motor allocation
| Period | Energy logic | Continuous motor power |
|---|
| Night (12 h) | battery 6 kWh usable, hold reserve for cloudy days | ≈100–150 W |
| Day (12 h) | solar surplus after hotel + partial battery cycling | ≈250–350 W |
| Burst (weather escape / maneuvers) | draw battery hard | 700–1,000 W+ |
Daily propulsion energy ≈ solar (3.4–4.5 kWh) − hotel (1.4 kWh) ≈ 2.0–3.0 kWh →
consistent with the allocations above.
6. Solar Panel Recommendation
- First choice: slim rigid monocrystalline panels with IEC 61701 salt-mist
certification, mounted on raised standoffs over the 2″ angle purlins so wave splash
drains through. Best $/W (~$0.5–0.8/W), best longevity, ~4.5 lb per 100 W.
- Second choice: ETFE-faced semi-flexible SunPower-cell panels (~$1–1.5/W,
~3 lb/100 W) glued directly — lighter but run hotter (−8–12% yield) and shorter lived.
- Either way: anodized frames, 316 SS fasteners with nylon isolation washers (aluminum
is galvanically fussy), and IP68 junctions (§10).
7. Speed Estimates
| Total motor power | Calm-water speed |
|---|
| 100 W | ≈2.7 kt |
| 150 W | ≈3.0 kt |
| 250 W | ≈3.5 kt |
| 350 W | ≈3.9 kt |
| 700 W | ≈4.9 kt |
| 1,000 W | ≈5.4 kt |
Wind correction (12–15 kt true wind, low flat silhouette, foil keels resist leeway):
| Heading | Night (100–150 W) | Day (250–350 W) |
|---|
| Into wind | ≈2.2–2.5 kt | ≈2.9–3.3 kt |
| Crosswind | ≈2.7–3.0 kt (minimal leeway) | ≈3.4–3.8 kt |
| Downwind | ≈3.2–3.5 kt | ≈4.2–4.6 kt |
Typical daily range: 70–90 nm → ~2,100–2,700 nm/month, entirely sun-powered in
Caribbean conditions. The legs-as-daggerboards mean crosswind legs cost almost nothing —
a genuine advantage of this layout for a wind-rode USV.
8. Capsize Risk & the "999 Days Out of 1000" Question
How to describe the tipping condition: this platform is extremely hard to capsize
statically (righting arms from widely spaced legs are enormous). The real danger is
dynamic: breaking/plunging crests whose height approaches the lateral leg spread —
roughly breaking waves >18–20 ft full-scale (>4.5–5 ft on the model) — or any sea
state (Hs ≥ ~3 m full / ~0.75 m model) containing frequent breaker faces slamming the deck
underside or solar array from below. Wind heel alone will never do it.
Is 999/1000 avoidance practical? Yes — with discipline:
- Hard operating rule: stay only where forecast Hs < 2 m (model: < 0.5 m),
checked against NOAA WAVEWATCH III / ECMWF / CMEMS, updated every 6–12 h.
- At 3–5 kt with 48–72 h of forecast lookahead, the drone can outrun or sidestep almost
any deteriorating area in the trades; squall lines are the main nuisance, and they're
visible on radar-based forecasts and (at night) by lightning sensors.
- Hurricane season (Jun–Nov): standing retrieval protocol — any named system within
1,000 nm or 5-day cone triggers recovery. This is the one non-negotiable rule.
- Residual risk after all that: unforecast rogue breakers, ship wake interactions,
vandalism/theft, entanglement. Accept ~<0.1%/day with conservative gates.
9. Foiling Analysis (Stabilizers as Hydrofoils)
Required lift = 2,500 N (562.5 lb). Lift = ½ρV²·S·C_L:
| Speed | Needed S·C_L | Verdict |
|---|
| 4.0 kt | 1.15 m² | Not attainable (foil stalls, C_L≤1.1) |
| 4.6 kt | 0.87 m² | Marginal with ~8 ft² of foil at C_L≈1.0 |
| 5.0 kt | 0.74 m² | Yes: ~8 ft² foil at C_L≈0.9 |
- Minimum takeoff speed ≈4.5–5 kt.
- Steady foiling power at 5 kt: induced + profile drag ≈240 N + residual leg drag ≈15 N
→ ~1.0–1.2 kW at the shafts.
- Endurance on a full battery: ≈5–6 h → ~28–33 nm. Useful as a weather-escape
sprint, not as efficient cruise — induced drag eats the drag savings. Net
efficiency vs. displacement mode is roughly neutral at best.
- Thruster placement: keep them at the leg trailing edges for now. Blowing the
foils (thrusters upstream of the stabilizer) buys maybe 10–20% extra lift at low speed via
slipstream, at the cost of interference drag — worth one A/B experiment on the model, but
don't redesign around it yet.
10. Thrusters, MTBF & Redundancy
M200 MTBF: Blue Robotics does not publish a formal MTBF for the M200 in
continuous seawater service (worth asking them directly for life-test data). Based on the
T-series track record and typical brushless-thruster reliability, a planning figure of
4,000–8,000 h MTBF in continuous use is defensible, with the dominant real failure
modes being biofouling/weed ingestion, fishing line, and connector corrosion rather than
motor windings.
Redundancy math (assume per-thruster MTBF = 6,000 h, 14-day mission, p(fail)=0.056,
independent failures):
- P(fewer than 2 usable thrusters on distinct legs) = P(≥5 failed) + P(exactly 2 left,
both on same leg) ≈ 3×10⁻⁵ per mission → loss-of-steerage "MTBF" ≈ millions of hours
statistically.
- In practice, common-mode causes (a weed event fouling several props, one flooded
connector pod, an ESC batch defect) dominate. Mitigations: reverse-purge prop cycles every
few hours, conformal-coated ESCs, connector inspection between missions.
- Your power architecture (each leg's pair fed only by its own battery/inverter) is the
right shape — keep the pairs electrically and physically isolated.
Alternatives considered: T200 (cheaper, but the enclosed shroud is exactly wrong
for sargasso), T500 (overkill/heavy), AliExpress ROV thrusters ($40–80, quality lottery),
small trolling motors (heavy, poor reverse). You're right that no affordable small RIM
drive exists today — they remain industrial/yacht toys. The open-prop M200 is the best
fit; add swept, weed-shedding blades and the purge routine and sargasso becomes manageable.
Vision-based avoidance (§12) is the second layer.
11. Stabilizer Tail Actuator & Spring-Pin Lock
Your "no sensor needed" reasoning is correct: setting the tail/elevator to a
fixed deflection commands a repeatable foil incidence (trim tab principle), so the wing's
angle is a pure function of tail angle + hydrodynamic hinge moment. Open-loop is fine.
Recommended hardware (per stabilizer)
| Function | Part | Approx. cost |
|---|
| Tail incidence | Waterproof metal-gear servo, 35–60 kg·cm (e.g., Hitec D-/HV class) with SS pushrod + horn | $30–45 |
| Lock pin | ¼″ stainless spring-indexing plunger (Vlier / Elesa / J.W. Winco style) in a fixed bracket | $8–15 |
| Detents | Hardened bushings / notched detent plates at lock positions on the wing root fitting | $10 |
| Unlock actuator | Small servo pulls the plunger against its spring (or a rotating cam wedge lifts the pin nose) | $15 |
| Hinges, brackets, 316 hardware | — | $20 |
| Total per station | | ≈$70–90 |
How it works exactly: the pin rides on its spring against a cam surface on the
root fitting. As waves rotate the wing toward neutral, the pin nose slides and
snap-into-detent when aligned — locked = wing pinned at zero incidence = effective
heave plate, as you intended. To unlock, the small servo retracts the plunger ~½″ against
the spring (fail-dangerous only if servo dies during unlock, which just means it
stays locked — the safe failure). Avoid solenoid pins: they corrode, overheat, and
fail-stroke. Skip the ball-lock pins (need perfectly axial alignment your waves won't
give you).
12. Salt Spray, Computing & Vision
Salt mitigation
- Solar wiring joints — your heat-shrink instinct is right, upgrade it: adhesive-lined
dual-wall 3:1 heat-shrink over every MC4/splice (plain "waterproof" MC4s wick eventually
under wave slam), plus dielectric grease inside, drip loops below every connection, and
potted/passivated OEM junction boxes. Inspect quarterly.
- Cameras: IP69K-rated units with replaceable glass fronts; hydrophobic nano-coating
(Rain-X-class) re-applied periodically; schedule rain/tender rinses; keep one spare dome
glass per camera. Sacrificial clarity beats pristine-until-fogged.
- Starlink Mini: IP56 — tolerates splash but not a bilge. Mount tilted ≥20° behind a
lipped polycarbonate shield with drain gaps; rinse routinely; budget one spare (they're
$599, and a dead dish strands the whole concept).
- General: sacrificial zinc anodes, no copper/graphite antifoulings on aluminum
(use aluminum-safe coatings), closed-cell gaskets everywhere, defined drain paths.
Computer
- Recommendation: Raspberry Pi CM4 with eMMC (you're right that eMMC beats SD for
corrosion/vibration mortality) + Hailo-8L M.2 accelerator (13 TOPS at ~1.5 W) for
onboard sargassum detection. If you want simpler integration, Pi 5 8GB + official AI Kit
works too — just manage its higher heat.
- Potting verdict: good idea, with two conditions. (1) Sylgard 184 conducts heat
poorly (~0.2 W/mK) — use a thermally conductive silicone potting instead
(MG Chemicals 832TC or similar, ~1–1.5 W/mK) and run a copper cold-plate from the SoC to
the external heatsink stack that pierces the potting into the water-cooled leg. (2) Pot the
connector cluster in a replaceable module so a cable failure doesn't scrap the computer.
At your duty cycle (throttled, ~8 W) this is very comfortable thermally.
- Competitors: Orange Pi 5 / Radxa ROCK 5 (RK3588, 6 TOPS NPU) are cheaper and
capable, but long-term supply/BSP polish trails Raspberry Pi; industrial SBCs (AAEON,
Compulab) buy you wider temp specs at 3–5× cost. For a potted, water-cooled, low-duty-cycle
node, CM4/Hailo is the sweet spot.
Night vision for sargasso
- 850 nm IR illuminator + no-IR-cut camera (Sony IMX462/585 sensor class) detects mat
lines to ~15–30 ft — adequate at 3 kt night speeds.
- Optional: FLIR Lepton 3.5 (~$200) — sargasso mats drift with the water and hold a
slightly different temperature; marginal but sometimes useful at dawn.
- Simplest robust policy: slow to ~2 kt at night and let the low speed itself be the
avoidance margin, with IR as confirmation.
13. Rescue Concept Review
(1) Upwind self-rescue — sound. Single-motor differential forward/reverse keeps
the nose home; the legs track like daggerboards while wind pushes hull. Reverse thrust is
~60–70% of forward, so budget asymmetric authority. Stabilizer-trim differential drag
steering as a backup is clever and costs nothing. Practice both before you need them.
(2) Auto-deploying backstay drogue — sound. Hinged, counterweighted flap: SS piano
hinge + HDPE/UHMW blade, ~$40. Add a light spring bias so it doesn't chatter at speed, and
make it removable for foiling tests. This one feature meaningfully raises survivability of
a comms-dead drone by keeping it nose-to-sea for human/drone pickup.
(3) Drone-to-drone rope hook — makes sense; suggested upgrades:
- Add a magnetic capture channel as primary (neodymium pot magnet recessed in the
funnel throat, steel target disc on the towed drone's bow-line float) with your rope/U-gate
as backup — magnets forgive aiming error far better than a thread-the-needle gate.
- Put a strobe + retroreflective tape + AIS target tag on the float so the rescue
drone finds it at night and the operator gets a bearing instantly.
- Tow at 1–2 kt with a weak link + quick release on the rescue bridle.
- For capsized targets: slow inverted towing works but abrades the solar array; a future
CO₂-inflated righting bag triggered by RF command is a cheap add-on worth prototyping.
- Print ArUco/AprilTag markers on the float now — your future onboard-AI hooking
will need exactly that fiducial, and it costs nothing today.
14. Competitive Landscape
| Vehicle | Size / Weight | Propulsion | Typical speed | Endurance / Range | Indicative price | Own code / instruments? | Self-righting? |
|---|
| Liquid Robotics Wave Glider SV3 | ~3 m float + subsurface glider / ~200–300 kg | Wave-energy glider + solar | 1.2–2 kt | 12+ months; trans-Pacific missions | ~$250k–500k (now largely sold as data services) | Payload bays via partners; vehicle software closed | Yes (float recovers) |
| Saildrone Explorer | 7 m / ~750 kg | 5.5 m rigid wing + solar | ~3 kt avg (0.5–10 kt) | 6–12 month missions | ~$300k-class (mission pricing dominates) | Payload bay + REST APIs for approved payloads | Yes (designed for 180° recovery) |
| AutoNaut 3.5/5.0 | 3.5–5 m / ~100–350 kg | Wave-adaptive foils + solar | 1.5–3.5 kt | Weeks–months | ~£150k–£250k | Research-friendly payload wells; commonly hosts customer sensors | Self-righting claimed via low CG |
| Sailbuoy (Offshore Sensing) | 2 m / ~60 kg | Wind sail + solar | ~1–2 kt | Months | ~$50k–100k | Modular payload bay, cooperative with researchers | Yes |
Why are they 10–30× your price?
- Engineering amortized for year-long unattended reliability, bespoke marine
components, low production volumes, certification/insurance, and — critically —
warranty + 24/7 ops support baked into price.
- Buyers (government, science) purchase guaranteed data, not vehicles.
Your positioning at 2× parts cost (~$22–26k)
You will not match a Wave Glider's 12-month autonomy or self-righting — and you don't need
to. Your wedge is expendability and swarm economics: at 1/10th–1/20th the price, an
operator who loses one to weather, weeds, or vandalism replaces it instead of mourning it.
For patrol, short-range sensing, education, and algorithm development, that's a different
(and unserved) product category. Be explicit in marketing: "disposable-class," operator-
in-the-loop recovery, no self-righting.
15. Market Estimate
- Fisheries & EEZ patrol: 100+ coastal states/territories with enforcement
shortages (your Anguilla DFMR use case generalizes across the Caribbean alone — dozens of
potential units).
- Science/monitoring: universities, NOAA-class agencies, coral-reef and Sargasso
Sea research, water-quality sensing — historically the anchor customers for small USVs.
- Aquaculture & ports: site patrols, security loops, bathymetry-lite.
- The broader USV market is measured in the hundreds of millions to low billions USD and
growing ~10–15%/yr; the sub-$50k patrol/data niche is small today (<$20–40M) but
nearly empty of competitors — first credible swarm-priced entrant defines it.
- Adjacent revenue: data subscriptions, "patrol-as-a-service," training/algorithms.
16. Cost Model — 5 Sets, Fabricated in China
| Item | Est. cost /set |
|---|
| Aluminum (legs, frame, heave plates, brackets) material + laser/form/weld | $2,800–3,500 |
| LiFePO₄ cells ~8 kWh + BMS | $1,200–1,500 |
| 6 × Blue Robotics M200 + ESCs | $2,400–3,000 |
| Solar ~850 W (lightweight, salt-rated) | $600–900 |
| Starlink Mini | $599 |
| CM4 + carrier + Hailo + cameras + AIS + LEDs + power conversion | $900–1,300 |
| Stabilizer servos/plungers/hardware | $150–200 |
| Rescue gear, wiring, fasteners, sealants, anodes | $300–400 |
| Shipping + customs (amortized) | $400–800 |
| Total per set (qty 5) | ≈$9,500–12,500 |
| At 2× parts = retail | ≈$19,000–25,000 |
17. Open Items / Risks Checklist
- Weight margin is ~1% — freeze the spec and weigh every assembly as built.
- Confirm M200 thrust/power curve and reverse-thrust % with Blue Robotics before finalizing
the control allocation matrix.
- Validate the drag model with a tow test or first sea trial; the speed table is ±30%.
- Leg venting/pressure relief for battery compartments; per-leg temp + leak sensors.
- Register the USV with Anguilla authorities; AIS + lighting keeps you COLREGS-defensible;
document patrol footage with GPS/time overlay if used as evidence.
- Decide final triangle size (11 vs 12 ft) once the weight audit is done — the extra
~150 W is worth 10 lb if you can find them.
```
**A few highlights worth calling out directly:**
1. **The weight budget closes, but barely** (~556 lb vs 562.5 target). I'd deliberately engineer to ~540 lb. Also note that at 562.5 lb you float ~5 in deeper than the scaled 27,500-lb waterline — same as the full-scale boat would, but decide consciously whether to accept that or trim to ~430–480 lb.
2. **The solar reality check:** ~850 W sounds like a lot, but after hotel load (~60 W) and Caribbean derating, your *sustainable* continuous motor power is only ~100–150 W at night and ~250–350 W by day — which still yields a very respectable 2.7–3.9 kt and ~2,500 nm/month.
3. **Foiling is a sprint tool, not an efficiency tool** — induced drag cancels the leg-drag savings. Treat it as the weather-escape gear, sized accordingly (~1.1 kW, ~30 nm range).
4. **On the M200 MTBF:** no public figure exists — I'd press Blue Robotics for life-test data, plan around 4,000–8,000 h, and remember that weeds and connectors will fail long before windings do.
Want me to generate a companion page with the control-allocation matrix for the six thrusters (including the single-motor-failure differential mode), or a parts spreadsheet version of the cost model?