Seastead Convoy Mode — Design Document

1. Convoy Mode Overview

Convoy mode enables a fleet of seasteads to travel together in a coordinated formation, maintaining precise relative positions on a grid. The system provides shared situational awareness, collective watch-keeping, mutual support, and the foundation for a mobile ocean community.

Core Capabilities

🎯 Formation Keeping

Each seastead maintains a precise position on a shared grid using RTK GPS and autopilot control of thrusters and stabilizers.

👁️ Shared Awareness

Cameras, radar, and AIS across all seasteads create a composite picture. Parallax ranging from multiple known positions gives precise distance measurements.

📡 Mesh Network

A high-bandwidth wireless mesh connects all seasteads, with directional antennas to neighbors and automatic multi-hop routing across the convoy.

🔄 Join/Leave

New seasteads can approach from outside and seamlessly join the convoy. Departing seasteads exit gracefully without disrupting formation.

👀 Night Watch

Distributed watch-keeping with human confirmation, AI monitoring, and shared duty schedules across the entire fleet.

🌊 Wave Sheltering

Multiple seasteads in formation may partially attenuate short-period wave energy for interior vessels through diffraction and energy absorption.

Key Assumptions

All seasteads run identical convoy software
All seasteads have Starlink internet, Class B AIS, and marine radar
Moving-base RTK GPS provides centimeter-level relative positioning
Each seastead has 6 rim-drive thrusters and 3 active stabilizers with triple-redundant power
Seasteads can maintain both position and heading via autopilot

2. Communications Mesh Network

2.1 Requirements Analysis

The convoy mesh network carries several categories of data between seasteads. Here is the bandwidth budget for a 100-seastead convoy:

Data Type	Per Seastead	Update Rate	Bandwidth
RTK GPS position + heading	200 bytes	10 Hz	2 kB/s
Thruster / stabilizer status	500 bytes	2 Hz	1 kB/s
Object sighting reports	1 kB	~1 Hz (event-driven)	1 kB/s
Compressed camera frames (for parallax)	50 kB	2 Hz	100 kB/s
Radar track data	5 kB	1 Hz	5 kB/s
AIS relay	300 bytes	0.03 Hz (every 30s)	10 B/s
Voice (VoIP, 1 active channel)	—	continuous	64 kbps
Watch confirmation / heartbeat	100 bytes	0.01 Hz (every 2 min)	<1 B/s
TOTAL per seastead (with headroom)			~2–5 Mbps

💡 Key Insight The per-seastead bandwidth requirement is modest — well under 10 Mbps. Even a single 20 MHz WiFi channel can handle this easily. The main challenge is range and reliability over open water, not raw throughput.

2.2 WiFi 5/6 on 5 GHz — Analysis

Why 5 GHz?

Less interference: Over open ocean there is virtually no competing 5 GHz traffic
More channels: 25+ non-overlapping 20 MHz channels vs. only 3 at 2.4 GHz
Higher throughput: Supports 80 MHz and 160 MHz channel widths for gigabit-class links
Directional antenna gain: Smaller antennas achieve the same gain at 5 GHz vs. 2.4 GHz

Propagation Over Water

Open water is an excellent RF environment — flat, no obstacles, and the Fresnel zone stays clear. The main limitation is Earth's curvature and Fresnel zone clearance. For antennas mounted at 3–5 m above sea level on the seastead walls/masts:

Parameter	WiFi 5 (802.11ac)	WiFi 6 (802.11ax)
Frequency	5.15 – 5.85 GHz	5.15 – 5.85 GHz (+ 6 GHz for 6E)
Max PHY rate (80 MHz, 2×2 MIMO)	867 Mbps	1,201 Mbps
Real-world throughput per link	200 – 400 Mbps	300 – 600 Mbps
Range with 23 dBi dish (over water, LOS)	5 – 15 km	5 – 15 km
Range with 15 dBi sector antenna	1 – 5 km	1 – 5 km
Latency	1 – 5 ms per hop	1 – 3 ms per hop
OFDMA (multi-user efficiency)	No	Yes — significant advantage for mesh

Range Over Water with Directional Antennas

For the Fresnel zone at 5.5 GHz, the 60% Fresnel radius at distance d is:

r_F = 0.6 × √(λ × d / 4) ≈ 0.6 × √(0.0545 × d / 4) ≈ 0.070 × √d (meters, with d in meters)

At 2 km: r_F ≈ 0.99 m. With antennas at 4 m above sea level, the Fresnel zone is clear out to ~3 km. Beyond that, some diffraction loss occurs but links remain viable. With 23 dBi dish antennas (Ubiquiti LiteBeam class), tested ranges of 10–15+ km over water are well-documented.

2.3 Hardware Recommendations

Budget Option

~$280

4× Ubiquiti NanoStation 5AC Loco (~$55 ea.)
13 dBi, 5 GHz, airMAX
Range: 3–8 km
Throughput: 100–200 Mbps
1× network switch ($30)
1× omni 2.4 GHz backup ($40)

Good for grid spacings up to ~2 km.

★ Recommended

~$550

4× Ubiquiti PowerBeam 5AC Gen2 (~$100 ea.)
25 dBi, 5 GHz, airMAX
Range: 10–25 km
Throughput: 200–400 Mbps
1× managed switch ($60)
1× omni 2.4 GHz backup ($40)

Excellent range/price ratio. Handles any reasonable convoy spacing.

High Performance

~$1,200

4× Ubiquiti airFiber 5XHD (~$250 ea.)
High gain, 5 GHz, custom protocol
Range: 10+ km
Throughput: 500+ Mbps
1× managed switch ($60)
1× omni 2.4 GHz backup ($40)

Overkill for most convoy sizes but future-proof.

Additional Communication Hardware

Item	Cost per Seastead	Purpose
LoRa 900 MHz module (e.g., RAK WisBlock)	$30 – $60	Emergency low-bandwidth backup. Range 10+ km, tiny data rate (~50 kbps). Works in bad weather when 5 GHz may degrade.
VHF marine radio (DSC capable)	$200 – $400	Standard maritime voice/distress. Required by maritime regulations.
Starlink Mini or Standard	$300 – $600 + $120/mo	Internet access for weather routing, marine traffic data, general comms.

2.4 Antenna Mounting Strategy

Your existing design includes a track around the top of the walls for a kite flying device. This track can serve double duty for antenna mounting. Here are the options:

Option A: Fixed Directional Antennas (Recommended)

If the autopilot maintains consistent heading (all seasteads face the same direction in formation), mount 4 directional antennas on fixed brackets at the midpoints of each wall side. Each antenna points toward the corresponding neighbor on the grid.

Simplest, cheapest, most reliable
Requires heading control (needed anyway for formation keeping)
If a seastead drifts off-heading by >30°, switch to the 2.4 GHz omni backup

Option B: Track-Mounted Antennas

Mount antennas on motorized trolleys that ride the wall-top track. As the seastead rotates, trolleys move to face the current neighbor positions. Uses the existing track infrastructure but adds motorized trolleys ($100–200 each) and control logic.

Option C: Omnidirectional Antennas (Simplest)

Use high-gain omnidirectional antennas (12–15 dBi, ~$80–150 each). No aiming required. Range is shorter (~1–3 km) but sufficient for close formations. Use 2–4 per seastead for redundancy.

2.5 Mesh Routing Software

The networking stack has three layers:

Layer	Technology	Notes
Physical / Link	Ubiquiti airMAX (TDMA) or standard 802.11ac/ax	airMAX is proprietary but well-proven for point-to-point. Standard WiFi works with commodity hardware.
Network / Routing	OSPF or Babel	OSPF is mature and widely supported. Babel is a modern mesh protocol that handles link quality and mobility well. Both run on Linux / OpenWrt.
Application	Custom convoy protocol over UDP multicast	Position broadcasts on a shared multicast group. Point-to-point TCP for file transfer and video.

✅ Practical Note Ubiquiti airMAX devices run airOS, which supports OSPF routing natively. For the recommended setup (4× PowerBeam 5AC), the configuration is: each PowerBeam forms a point-to-point bridge to its neighbor, OSPF runs across all four links, and application data is multicast over UDP. This is a well-understood configuration used in WISP (wireless ISP) deployments worldwide.

2.6 Data Flow Architecture


    ┌───────────────────────────────────────────────────────────┐
    │                    STARLINK (Internet)                     │
    │          Weather routing, MarineTraffic, backups           │
    └────────────────────────┬──────────────────────────────────┘
                             │
    ┌────────────────────────▼──────────────────────────────────┐
    │                  LOCAL GATEWAY / ROUTER                    │
    │            (MikroTik hAP ac³ or Ubiquiti EdgeRouter)      │
    ├──────────┬──────────┬──────────┬──────────┬───────────────┤
    │  WiFi    │  WiFi    │  WiFi    │  WiFi    │  Local AP     │
    │  Link N  │  Link E  │  Link S  │  Link W  │  (on-board)   │
    │  (25dBi) │  (25dBi) │  (25dBi) │  (25dBi) │  (omni)       │
    │  5 GHz   │  5 GHz   │  5 GHz   │  5 GHz   │  2.4/5 GHz    │
    └────┬─────┴────┬─────┴────┬─────┴────┬─────┴───────────────┘
         │          │          │          │
    ┌────▼──────────▼──────────▼──────────▼─────────────────────┐
    │              OSPF / BABEL MESH ROUTING                     │
    │         Automatic path selection, failover                 │
    ├───────────────────────────────────────────────────────────┤
    │           CONVOY APPLICATION (UDP multicast)               │
    │  ┌─────────────┬──────────────┬─────────────┐             │
    │  │Position DB   │ Object Tracker│ Watch Mgr   │             │
    │  │(all vessels) │ (all targets) │ (duty log)  │             │
    │  └─────────────┴──────────────┴─────────────┘             │
    └───────────────────────────────────────────────────────────┘

3. Navigation & Positioning

3.1 Moving-Base RTK GPS

Standard GPS gives ~2–5 m accuracy. For formation keeping with seasteads that are 44 feet (13.4 m) across, we need centimeter-level relative accuracy. Moving-base RTK achieves this by having each seastead share its raw GNSS carrier-phase observations with all other seasteads via the mesh network.

How It Works

Each seastead has an RTK-grade GNSS receiver (e.g., u-blox ZED-F9P) with a survey-grade antenna
Each receiver outputs raw observation data (pseudorange + carrier phase) at 10+ Hz
These observations are broadcast over the mesh network
Each seastead computes relative positions to every other seastead by processing differential carrier-phase measurements
Result: 1–2 cm relative accuracy between any two seasteads in the convoy

💡 Why "Moving Base"? Traditional RTK uses a fixed base station on land. In a convoy, every seastead is moving — there is no fixed reference. The "moving base" technique treats one seastead (or the convoy centroid) as a virtual base. Since all seasteads share the same satellite constellation and atmospheric conditions, the common-mode errors cancel out beautifully. The relative position accuracy is independent of absolute position accuracy.

Hardware Options

Solution	Cost	Accuracy (relative)	Update Rate	Notes
u-blox ZED-F9P + good antenna	$200 – $350	1 – 2 cm	10 – 20 Hz	Best value. Requires RTK software (RTKLIB, STRSVR).
ArduSimple RTK2Go	$300 – $400	1 – 2 cm	10 Hz	Pre-assembled ZED-F9P board. Easy to integrate.
Emlid Reach RS2+	$2,500	1 cm + 1 ppm	10 Hz	All-in-one with built-in radio, battery, and software. Very polished.
Septentrio mosaic-X5	$1,500 – $2,000	1 cm	100 Hz	Professional grade. Multi-constellation (GPS, GLONASS, Galileo, BeiDou).

Recommended: u-blox ZED-F9P

At ~$250 per seastead (module + antenna), the ZED-F9P offers an unbeatable price/performance ratio. It supports GPS, GLONASS, Galileo, and BeiDou simultaneously, giving excellent satellite coverage worldwide. The open-source RTKLIB software handles the differential processing. Run it on a Raspberry Pi 4/5 or similar single-board computer.

3.2 Heading Determination

In addition to position, each seastead needs accurate heading (yaw) to maintain orientation on the grid. Two approaches:

Dual-antenna GPS: Mount two GNSS antennas ~2 m apart along the seastead's fore-aft axis. The baseline between them gives heading accurate to ~0.5°. Uses the same ZED-F9P (it supports dual-antenna mode).
RTK GPS + IMU fusion: Use a 9-axis IMU (e.g., BNO085, ~$30) fused with RTK GPS in a Kalman filter. This gives smooth, high-rate (100+ Hz) heading estimates.

Recommendation: Use both. The dual-GPS antenna gives absolute heading; the IMU gives high-rate updates between GPS epochs. Fuse them with a complementary or Kalman filter.

3.3 Coordinate System

The convoy uses a local East-North-Up (ENU) coordinate frame:

Origin: The convoy leader's position (or the centroid of all seasteads)
East/North: Local tangent plane
Grid positions: Defined as offsets (ΔE, ΔN) from the origin in meters

Each seastead's target grid position is a fixed (ΔE, ΔN) offset from the convoy origin. As the convoy moves, the origin moves; the relative offsets stay constant.

4. Formation Keeping & Autopilot

4.1 Grid Design

The convoy arranges seasteads on a rectangular grid. The grid spacing is a key design parameter that affects safety (collision avoidance), communications (antenna alignment), and wave interaction.

Grid Spacing	Safety Margin	Comms Ease	Wave Sheltering	Walkway Transfer
50 m (164 ft)	Tight — requires precise control	Easy — omni antennas work	Best — closest formation	Needs intermediate boat
100 m (328 ft)	Comfortable	Easy — small directional antennas	Moderate	Needs dinghy
200 m (656 ft)	Generous	Easy — directional antennas	Modest	Needs dinghy
500 m (1640 ft)	Very safe	Needs directional antennas	Minimal for individual rows	Needs dinghy

💡 Recommendation Start with 100 m grid spacing. This gives a comfortable safety margin, easy communications, and the convoy can be reconfigured to tighter spacing (50 m) when conditions permit and the group agrees. The grid spacing should be a runtime parameter, adjustable by the convoy leader.

4.2 Autopilot Control Loop

Each seastead runs a position-keeping autopilot that drives its 6 rim-drive thrusters and 3 active stabilizers:


    ┌──────────────────────────────────────────────────────────────┐
    │                    AUTOPILOT CONTROL LOOP                     │
    │                                                              │
    │   Target Position (ΔE, ΔN, ψ)     ← From convoy manager     │
    │         │                                                    │
    │         ▼                                                    │
    │   ┌─────────────┐     ┌──────────────┐                      │
    │   │  Position    │────▶│  Heading     │                      │
    │   │  Controller  │     │  Controller  │                      │
    │   │  (PID/MPC)   │     │  (PID)       │                      │
    │   └──────┬──────┘     └──────┬───────┘                      │
    │          │                   │                               │
    │          ▼                   ▼                               │
    │   ┌──────────────────────────────────┐                      │
    │   │   Thruster Allocation            │                      │
    │   │   (6 rim-drives: surge, sway,    │                      │
    │   │    yaw control)                  │                      │
    │   └──────────────────────────────────┘                      │
    │          │                                                   │
    │          ▼                                                   │
    │   ┌──────────────────────────────────┐                      │
    │   │   Stabilizer Allocation          │                      │
    │   │   (3 active foils: pitch, roll   │                      │
    │   │    damping)                      │                      │
    │   └──────────────────────────────────┘                      │
    │                                                              │
    │   Current Position/Heading ← RTK GPS + IMU                  │
    │   Current State          ← Thruster RPMs, battery SOC, etc. │
    └──────────────────────────────────────────────────────────────┘

Control Approach

Outer loop (1 Hz): Position error → desired force/torque vector. Uses PID or Model Predictive Control (MPC). The MPC approach can anticipate waves and current for smoother station-keeping.
Inner loop (10–50 Hz): Desired force/torque → individual thruster commands. Uses a thruster allocation algorithm that distributes thrust across the 6 rim-drives, respecting limits and optimizing for efficiency.
Stabilizer control (10 Hz): IMU pitch/roll rate → elevator servo commands on the 3 active stabilizer foils. This damps wave-induced motion independently of position keeping.

4.3 Coordination Between Seasteads

In convoy mode, seasteads need to coordinate their thrust to avoid creating wakes or currents that disturb neighbors. Two approaches:

Independent control (simple): Each seastead maintains its own position independently. Works well if spacing is large enough that hydrodynamic interaction is small.
Coordinated control (advanced): Seasteads share their thrust commands, and a central coordinator optimizes the fleet's collective thrust to minimize wake interaction and energy use. More complex but more efficient for close formations.

Recommendation: Start with independent control. Add coordination later if the fleet grows large or spacing decreases below 50 m.

4.4 Failure Modes

The triple-redundant power system (each leg has its own battery, charge controller, and inverter) means a single failure doesn't disable the seastead. But in convoy mode, additional failure modes exist:

Failure	Detection	Response
RTK GPS loss	Position fix degrades; HDOP increases	Fallback to standard GPS + IMU dead reckoning. Increase grid spacing. Alert convoy.
Mesh link loss (one direction)	OSPF detects link down	Traffic reroutes through other links. No immediate action needed if other links are healthy.
Complete comms loss	No heartbeat received by neighbors	Seastead maintains last known course/speed. Convoy broadcasts alerts on AIS and VHF. After timeout (5 min), seastead enters "safe mode": hold position, flash lights.
Thruster failure (one)	Motor controller reports error	Thruster allocation redistributes thrust to remaining 5 drives. Slight performance reduction.
Stabilizer failure (one)	Actuator feedback	Disable that stabilizer. Remaining 2 provide reduced but adequate roll/pitch damping.
Battery / power failure (one leg)	Voltage monitoring	Isolate failed leg. Two remaining legs power all systems at reduced capacity. Convoy can tow or escort the disabled seastead.

5. Situational Awareness & Tracking

5.1 Multi-Layer Detection

The convoy uses four complementary detection layers. Each seastead contributes to the collective awareness, creating a picture far superior to what any single vessel could achieve.

Layer 1: AIS

Range: 20–40 NM (VHF line-of-sight)
Detects: All Class A & B AIS-equipped vessels
Data: MMSI, position, course, speed, vessel name/type
Cost: $300–600 per seastead (Class B transponder)

Limitation: Small boats, fishing vessels, and military ships may not broadcast AIS.

Layer 2: Radar

Range: 2–16 NM (depends on antenna height and target size)
Detects: All solid objects (boats, debris, land, weather)
Data: Range, bearing, radar cross-section
Cost: $1,000 – $3,000 (compact solid-state radar)

Limitation: No identification; false returns from waves and rain.

Layer 3: Cameras + AI

Range: 0.5 – 5 NM (depends on conditions and optics)
Detects: Visual targets, navigation lights, hull shapes
Data: Bearing, classification, imagery
Cost: $500 – $2,000 (4× PTZ cameras + GPU/computer)

Advantage: Can identify vessel type and read names. Works at night with IR/low-light cameras.

Layer 4: Parallax Ranging

Range: 1 – 10+ NM (depends on baseline and camera resolution)
Detects: Any target visible to 2+ seasteads
Data: Precise distance via triangulation
Cost: "Free" — uses cameras already present + RTK GPS positions

Unique advantage: No radar emissions, passive and covert. Accuracy improves with larger baseline (more seasteads).

5.2 Parallax Ranging — Detailed Analysis

When two or more seasteads at precisely known positions (from RTK GPS) detect the same object in their cameras, the distance to the object can be computed by triangulation:

d = B / tan(θ₁ - θ₂)

Where B is the baseline (distance between the two seasteads), and θ₁, θ₂ are the bearing angles from each seastead to the target.

Accuracy Estimates

Baseline (B)	Target Range (d)	Angular Resolution	Range Accuracy (Δd)
100 m	1 km	0.1° (camera)	±17 m
500 m	1 km	0.1°	±3.5 m
500 m	5 km	0.1°	±87 m
2 km	5 km	0.1°	±22 m
2 km	10 km	0.1°	±87 m
5 km	10 km	0.1°	±35 m

With multiple seasteads triangulating the same target (3–5 seasteads in a line for a beam-on target), the accuracy improves further through least-squares estimation. Sub-degree camera resolution is achievable with telephoto lenses and image processing.

✅ Synergy with Radar Parallax ranging is passive (no emissions), works on visual identification, and improves with fleet size. Combined with radar (which gives precise range but limited identification) and AIS (which gives identification but relies on the target cooperating), the convoy achieves very robust tracking.

5.3 Unified Object Tracking Database

All detection data from all seasteads feeds into a shared, distributed object tracking database. Each tracked object has:

Unique track ID (assigned by the first seastead to detect it)
Position and velocity (constantly updated from all sources)
Classification (cargo ship, fishing boat, sailboat, debris, unknown)
Confidence level (increases with multiple detections)
Source list (which seasteads are contributing to this track)
AIS correlation (matched to AIS MMSI if available)

The tracking algorithm uses a multi-hypothesis tracker (MHT) or probabilistic data association (PDA) to handle false alarms and track splitting/merging. This is well-established technology in maritime surveillance.

5.4 Night Watch — AI Augmentation

During night watch, human operators are augmented by AI systems running on each seastead:

Camera AI: YOLO-v8 or similar neural network continuously classifies objects in camera feeds. Detects navigation lights, hulls, and wakes. Runs on a local GPU (NVIDIA Jetson Nano/Orin, ~$200–500).
Radar AI: Automatic target tracking with track-while-scan. Identifies CPA (closest point of approach) and alerts on collision risk.
Fusion engine: Combines all sensor data into a unified picture. Uses Kalman filtering for track smoothing and prediction.
Alert escalation: AI generates alerts that are sent to the human watch-keeper. If no human confirmation within 30 seconds, alerts escalate to the convoy.

6. Watch & Safety Protocols

6.1 Watch System

Each seastead runs a standard watch rotation (e.g., 4 hours on, 8 hours off). The convoy can share the load — not every seastead needs a human watch-keeper at all times, but a minimum number must be on duty across the fleet.

Watch Confirmation Protocol

Every 10 minutes, the watch system prompts the watch-keeper for confirmation (audible alarm + screen prompt).
The watch-keeper presses a button or responds via touchscreen within 60 seconds.
The confirmation (timestamp + seastead ID) is broadcast to the convoy.
If no confirmation is received within 60 seconds, the system:
- Escalates the alert to the seastead (louder alarm, wake sleeping crew)
- Notifies the convoy after 2 more minutes
- The convoy can dispatch assistance or adjust the watch schedule

Convoy Watch Dashboard

Every seastead displays a shared dashboard showing:

🟢 Green: Watch-keeper confirmed in last 10 minutes
🟡 Yellow: Watch-keeper overdue (10–15 minutes)
🔴 Red: Watch-keeper absent (>15 minutes), AI-only monitoring
⚪ Gray: Seastead not required to have watch (rest period)

Minimum Watch Requirement

The convoy should maintain a minimum number of active watch-keepers at all times. For a fleet of N seasteads:

Min watch-keepers = max(2, ceil(N × 0.2))

For a 10-seastead convoy: at least 2 seasteads must have confirmed human watch-keepers at all times. The remaining 8 can be in AI-only mode during their rest periods. This dramatically reduces fatigue compared to single-vessel operations.

6.2 Consensus Protocol

Certain decisions require convoy-wide agreement. A lightweight consensus protocol handles these:

Decision	Authority	Mechanism
Course/speed changes	Convoy leader	Leader decides after consulting. Others follow.
Grid spacing changes	Convoy leader	Leader proposes; fleet acknowledges.
Join/leave requests	Convoy leader + 1	Leader approves; at least one other seastead confirms.
Emergency (any seastead)	Any seastead	Any seastead can trigger "all stop" or "emergency maneuver." Others comply immediately. Post-hoc review.
Leader election	All seasteads	Raft consensus algorithm. Leader rotates every 24 hours or on failure.
Watch schedule	Convoy leader	Leader proposes schedule based on crew availability. Others confirm.

Leader Election (Raft)

The convoy uses the Raft consensus algorithm to elect and maintain a leader. Raft is well-understood, simple to implement, and proven in production distributed systems. Key properties:

One leader at a time; all decisions flow through the leader
If the leader's seastead goes offline, a new leader is elected within seconds
Leader role rotates every 24 hours to share the workload
Any seastead can propose itself as leader candidate

6.3 Emergency Procedures

Man Overboard (MOB)

Any seastead detects MOB (via person-tracking AI, watch alarm, or manual report)
MOB alert broadcast to entire convoy immediately
All seasteads mark the GPS position of the MOB
Nearest seastead(s) break formation to assist
Convoy slows and holds position
Dinghy can be deployed from the nearest seastead

Seastead Disabled

Disabled seastead broadcasts "disabled" status with position
Convoy adjusts formation to leave a gap
Adjacent seasteads offer assistance (power, towing)
Two seasteads can connect with a walkway for crew transfer
If towing is needed, the walkway connections are rated for some towing force

Weather Emergency

Weather routing AI (via Starlink) detects approaching storm
Convoy leader receives recommendation: scatter, heave-to, or run
Leader decides; all seasteads execute simultaneously
Grid spacing increases automatically as seas build
After storm passes, convoy regroups at a designated rendezvous point

7. Wave Interaction Analysis

7.1 The Question

Could a large number of seasteads traveling together in formation reduce the average wave height experienced by interior seasteads? Each seastead's three SWATH legs interact with the wave field through diffraction, scattering, and energy absorption. In a large convoy, these interactions might combine to provide measurable wave sheltering.

7.2 Single Seastead in Waves

Each seastead has three NACA 0030 foil-shaped legs, oriented vertically with the chord (8.5 ft / 2.59 m) horizontal and the span (14.5 ft / 4.42 m) vertical. Half the span is submerged (draft = 2.21 m / 7.25 ft).

Key Dimensions at the Waterline

Parameter	Value	Notes
Waterplane width (beam seas)	2.59 m (8.5 ft)	Full chord presents to side-on waves
Waterplane width (head seas)	0.78 m (2.55 ft)	Only thickness presents to head-on waves
Draft	2.21 m (7.25 ft)	50% of 14.5 ft span
Total waterplane area (3 legs)	~4.2 m² (beam seas)	Very small — classic SWATH advantage
Submerged volume (3 legs)	~14 m³	NACA 0030 cross-section × 3 legs × draft

The Diffraction Parameter (ka)

The key parameter governing wave-body interaction is ka = 2πa/λ, where a is the characteristic body dimension and λ is the wavelength.

ka < 0.5 → waves dominate, weak scattering (long waves relative to body)
ka ≈ 1.0 → comparable scales, strong interaction
ka > 2.0 → body dominates, strong shadow (short waves relative to body)

Wave Period (T)	Wavelength (λ)	ka (a = 1.3 m)	Regime	Scattering Strength
2 s	6.2 m	1.31	Strong interaction	■■■■■
3 s	14.1 m	0.58	Moderate interaction	■■■■□
4 s	25.0 m	0.33	Moderate interaction	■■■□□
5 s	39.0 m	0.21	Weak interaction	■■□□□
6 s	56.2 m	0.15	Weak interaction	■■□□□
8 s	100 m	0.08	Very weak	■□□□□
10 s	156 m	0.05	Negligible	■□□□□
12 s	225 m	0.04	Negligible	□□□□□
15 s	351 m	0.02	Negligible	□□□□□

⚠️ Key Finding The SWATH legs are small relative to typical ocean waves. For the dominant swell periods (8–15 s), ka < 0.1, meaning the legs barely disturb the wave field. Wave sheltering effects will be concentrated in the short-period wind chop (T < 6 s), not the long swells.

7.3 Wave Shadow Behind a Single Leg

Behind a single obstacle in water waves, there is a "shadow zone" where wave amplitude is reduced. The shadow depth (amplitude reduction) depends on ka and distance behind the obstacle.

For a vertical cylinder of radius a, the wave amplitude at distance r directly behind it is approximately:

A(r)/A₀ ≈ 1 − (2/π) × arctan(1/(2ka)) × √(a/r)

Applying this to one NACA 0030 leg (a ≈ 1.3 m) at 30 m behind:

Wave Period	ka	Amplitude at 30 m behind	Reduction
3 s	0.58	84%	16%
5 s	0.21	85%	15%
8 s	0.08	86%	14%
10 s	0.05	87%	13%

Note: This simple formula gives a weak dependence on ka for the far-field shadow, which suggests that even for long waves there is some shadow effect. However, this formula is approximate and doesn't capture the full diffraction physics. In reality, for ka < 0.1, the wave essentially "wraps around" the obstacle with minimal net shadowing beyond a few body widths. The actual reduction for long waves would be much smaller than these estimates suggest.

7.4 Array Effect — A Row of Seasteads

A row of seasteads perpendicular to the wave direction creates a periodic array of obstacles. The wave transmission through this array depends on the blockage ratio and the spacing-to-wavelength ratio.

Blockage Parameters

For a row of seasteads with spacing D between centers, the blockage ratio is:

Blockage ratio = (3 legs × chord width) / D = (3 × 2.59 m) / D = 7.77 / D

Grid Spacing (D)	Blockage Ratio	Gap Fraction
50 m	15.5%	84.5%
100 m	7.8%	92.2%
200 m	3.9%	96.1%

Bragg Resonance

For a periodic array of obstacles, there is a Bragg-like resonance condition where the array becomes strongly reflective. This occurs when:

2D ≈ nλ (n = 1, 2, 3, ...)

At resonance, the scattered waves from each obstacle reinforce each other, creating strong wave attenuation in the transmitted direction.

Grid Spacing (D)	Bragg Resonant λ (2D = λ)	Corresponding Wave Period	Common Sea State?
50 m	100 m	8.0 s	Yes — typical wind waves / moderate swell
100 m	200 m	11.3 s	Yes — common swell period
200 m	400 m	16.0 s	Less common — long swell

💡 Bragg Resonance — Design Opportunity By choosing the grid spacing, you can position the Bragg resonance at the dominant wave period of your operating area. For example, if you're sailing in an area with dominant 8-second wind waves, a 50 m grid spacing would put the Bragg resonance right at 8 seconds, maximizing wave attenuation for those waves. However, the strength of the Bragg effect depends on the individual scatterer's reflection coefficient, which is weak for the SWATH legs.

7.5 Multi-Row Cumulative Attenuation

For a convoy with multiple rows of seasteads, the wave attenuation accumulates. If each row reduces the wave amplitude by fraction f, after N rows:

A_N / A₀ = (1 − f)^N

Estimated Attenuation per Row

Based on the diffraction analysis and the small blockage ratio, here are conservative estimates of wave amplitude attenuation per row for a 100 m grid spacing:

Wave Period	Type	Est. Attenuation per Row	After 5 Rows	After 10 Rows	After 20 Rows
2 – 3 s	Wind chop	10 – 20%	40 – 65% reduction	65 – 88% reduction	88 – 99% reduction
4 – 5 s	Short wind waves	5 – 12%	23 – 47% reduction	41 – 72% reduction	65 – 92% reduction
6 – 8 s	Wind waves	3 – 7%	14 – 31% reduction	26 – 52% reduction	46 – 77% reduction
10 – 12 s	Swell	1 – 3%	5 – 14% reduction	10 – 26% reduction	18 – 46% reduction
15+ s	Long swell	< 1%	< 5% reduction	< 10% reduction	< 18% reduction

⚠️ Important Caveats

These are order-of-magnitude estimates based on simplified diffraction theory. A proper analysis requires numerical simulation (BEM/WAMIT/AQWA).
The actual attenuation depends on wave direction, sea state spectrum, and the specific hydrodynamic response of the SWATH legs.
The SWATH legs are very "transparent" to long waves. The primary benefit is for short-period chop, not ocean swell.
Wave energy is conserved — it's redistributed, not destroyed. The reflected waves may increase wave heights outside the convoy.

7.6 Additional Wave-Related Effects

Wave Energy Absorption by SWATH Legs

When the seastead heaves, pitches, or rolls in response to waves, the submerged SWATH foils create radiation waves that radiate energy away from the seastead. This energy comes from the seastead's own motion — effectively, the seastead absorbs wave energy and converts it to radiated waves that propagate away. The radiation damping of SWATH foils is significant due to their large submerged area.

In a convoy, this means each seastead is acting as a passive wave energy absorber, slightly reducing the total wave energy within the convoy over time. The effect is small per seastead but could be meaningful for a large fleet.

Wind Sheltering

The 7-foot walls of the living area create a wind shadow on the leeward side. Since local wind waves depend on the wind speed over the water surface, reducing the wind speed within the convoy directly reduces locally-generated wind wave height.

Wind wave height H ∝ U² (quadratic with wind speed)
A 30% reduction in wind speed → ~50% reduction in locally-generated wave height
The sheltering zone extends roughly 10–20× the wall height downwind (70–140 ft / 20–43 m)
For a tight formation (50 m spacing), interior seasteads would experience significantly reduced wind, and thus reduced locally-generated chop

Wake Interaction

When the convoy is underway, each seastead generates a Kelvin wake. However, SWATH vessels have very low wave-making resistance due to their small waterplane area, so the wakes are much smaller than conventional ships. At convoy speeds (3–5 knots), the wake amplitude would be minimal.

7.7 Summary — Wave Interaction

✓ What Works

Short-period chop (T < 5s): Measurable attenuation, especially for deep convoys (10+ rows)
Wind sheltering: Significant reduction in locally-generated waves for tight formations
Low wake generation: SWATHs don't disturb each other like conventional ships
Bragg resonance: Grid spacing can be tuned to target specific wave periods

✗ What Doesn't

Long-period swell (T > 10s): Legs are too small to significantly attenuate
Small convoys (fewer than 5 seasteads): Cumulative effect is negligible
Head seas: Thinnest part of foil presents to oncoming waves

✅ Bottom Line For a large convoy (20+ seasteads) in moderate seas, interior seasteads would experience noticeably reduced short-period wave energy — perhaps 50%+ reduction in wind chop amplitude. The effect on long-period swell is minimal. The primary comfort benefit of convoy mode comes from the SWATH design itself (low wave response) rather than fleet interaction. The wave sheltering is a real but secondary benefit that improves with fleet size.

8. Joining & Leaving the Convoy

8.1 Joining Protocol

A new seastead approaching the convoy follows a structured protocol:

Figure 1: A new seastead approaches the convoy from outside and is assigned a target grid position.

Step-by-Step Joining Process

Request: The approaching seastead contacts the convoy leader via Starlink or VHF, requesting to join. It provides its position, heading, speed, and vessel info.
Assignment: The convoy leader's software assigns an available grid position (e.g., S12 in the diagram). The position is chosen to be on the edge of the convoy, minimizing disruption.
Approach: The new seastead navigates to within one grid spacing (100 m) of its assigned position, approaching from outside the convoy. It maintains its own navigation.
Convoy Lock: When the new seastead is within half a grid spacing (50 m) of its assigned position AND its relative velocity is below a threshold (e.g., 1 knot):
- The system announces: "Convoy mode activated for [Seastead ID]"
- The new seastead's autopilot switches to convoy formation-keeping mode
- It begins broadcasting on the convoy mesh network
- All other seasteads acknowledge the new member
Verification: After 5 minutes of stable position-keeping, the new seastead is confirmed as a full convoy member and added to the watch rotation.

8.2 Leaving Protocol

Request: The departing seastead informs the convoy leader that it wishes to leave. It specifies the intended direction of departure.
Approval: The convoy leader confirms and clears the departure vector (ensures no other seasteads are in the way).
Departure: The seastead transitions from convoy mode to independent navigation, smoothly moving out of formation.
Grid reorganization: If desired, remaining seasteads can shift positions to fill the gap. This is optional — leaving a gap is also fine.

8.3 Walkway Connection (Side-by-Side)

When two seasteads need to be connected with a walkway (for crew transfer, maintenance, or community gathering), they move to adjacent grid positions and reduce their spacing to ~15–20 feet. The walkway is deployed between them.

In this configuration, both seasteads' autopilots cooperate to minimize relative motion, particularly when someone is on the walkway. The system:

Monitors walkway occupancy (motion sensor or manual button)
When occupied, both autopilots switch to "walkway mode" — tighter position control, prioritizing minimization of relative sway and heave
Active stabilizers on both seasteads coordinate to dampen relative roll
Maximum allowable relative motion is monitored; alerts if exceeded

9. System Architecture

9.1 Per-Seastead Software Stack


┌─────────────────────────────────────────────────────────────┐
│                     USER INTERFACE                           │
│  Dashboard · Map · Watch Status · Alerts · Configuration    │
├─────────────────────────────────────────────────────────────┤
│                    CONVOY APPLICATION                        │
│  ┌───────────────┐ ┌──────────────┐ ┌─────────────────────┐ │
│  │ Formation      │ │ Watch        │ │ Object Tracking     │ │
│  │ Manager        │ │ Manager      │ │ Database            │ │
│  │ (grid pos,     │ │ (rotation,   │ │ (AIS + radar +      │ │
│  │  join/leave,   │ │  confirm,    │ │  camera fusion,     │ │
│  │  leader elect) │ │  alerts)     │ │  parallax ranging)  │ │
│  └───────┬───────┘ └──────┬──────┘ └──────────┬──────────┘ │
├──────────┼────────────────┼────────────────────┼────────────┤
│          │     CONVOY COMMUNICATIONS LAYER      │            │
│          │  ┌─────────────────────────────────┐ │            │
│          │  │ Position broadcast (UDP mcast)   │ │            │
│          │  │ Object database sync (TCP)       │ │            │
│          │  │ Watch/consensus messages (UDP)   │ │            │
│          │  │ Voice/video (VoIP/WebRTC)        │ │            │
│          │  └────────────┬────────────────────┘ │            │
├──────────┼───────────────┼──────────────────────┼────────────┤
│          │      MESH NETWORK (OSPF/Babel)        │            │
│          │  ┌─────────────────────────────────┐ │            │
│          │  │ 4× Directional WiFi 5 GHz links │ │            │
│          │  │ 1× Omni 2.4 GHz backup          │ │            │
│          │  │ LoRa 900 MHz emergency          │ │            │
│          │  │ Starlink internet gateway        │ │            │
│          │  └────────────┬────────────────────┘ │            │
├──────────┼───────────────┼──────────────────────┼────────────┤
│          │       AUTOPILOT / CONTROL             │            │
│  ┌───────┴───────┐ ┌────┴────────┐ ┌───────────┴──────────┐ │
│  │ Position       │ │ Thruster    │ │ Stabilizer           │ │
│  │ Controller     │ │ Allocator   │ │ Controller           │ │
│  │ (PID/MPC)      │ │ (6 drives)  │ │ (3 active foils)     │ │
│  └───────┬───────┘ └────┬────────┘ └───────────┬──────────┘ │
├──────────┼───────────────┼──────────────────────┼────────────┤
│                    SENSOR FUSION                              │
│  ┌──────────┐ ┌──────────┐ ┌──────────┐ ┌─────────────────┐ │
│  │ RTK GPS  │ │ IMU/Compass│ │ Radar    │ │ Cameras (4×PTZ) │ │
│  │ (2 cm)   │ │ (100 Hz)  │ │ (24 NM)  │ │ + AI detection  │ │
│  └──────────┘ └──────────┘ └──────────┘ └─────────────────┘ │
├─────────────────────────────────────────────────────────────┤
│                    HARDWARE LAYER                            │
│  6× Rim-drive thrusters · 3× Active stabilizers · Batteries │
│  Solar · 3× Charge controllers · 3× Inverters               │
└─────────────────────────────────────────────────────────────┘

9.2 Computing Hardware

Component	Recommended	Cost	Purpose
Main computer	Raspberry Pi 5 (8 GB) or NVIDIA Jetson Orin Nano	$80 – $500	Runs autopilot, convoy software, mesh routing
AI vision processor	NVIDIA Jetson Orin NX	$500 – $700	Camera AI inference (object detection, classification)
RTK GPS	u-blox ZED-F9P + antenna	$250	2 cm positioning
IMU	BNO085 or VectorNav VN-100	$30 – $300	Heading, pitch, roll at 100+ Hz
Network switch	MikroTik CRS305 or similar managed switch	$60 – $150	Connects all network interfaces
UPS / power conditioning	12V/24V to 5V converters + battery backup	$50 – $100	Reliable power for electronics

9.3 Data Communication Protocol

The convoy application uses a lightweight protocol over UDP multicast for real-time data and TCP for reliable transfers:

UDP Multicast Messages (Real-Time)


// Position broadcast — every seastead, 10 Hz
struct PositionMsg {
    uint32_t  seastead_id;
    uint64_t  timestamp_us;    // GPS time in microseconds
    double    lat;             // Latitude (degrees)
    double    lon;             // Longitude (degrees)
    float     heading;         // True heading (degrees)
    float     speed_kn;        // Speed over ground (knots)
    float     pitch_deg;       // Pitch angle
    float     roll_deg;        // Roll angle
    float     battery_soc;     // Battery state of charge (0–1)
    uint8_t   watch_status;    // 0=off, 1=on, 2=confirmed
    uint8_t   thruster_health; // Bitmask: 6 bits for 6 thrusters
};

// Object sighting report — event-driven
struct ObjectSightMsg {
    uint32_t  source_id;       // Reporting seastead
    uint32_t  track_id;        // Unique track ID
    uint64_t  timestamp_us;
    double    lat;
    double    lon;
    float     course_deg;
    float     speed_kn;
    uint8_t   classification;  // AIS_A, AIS_B, radar, visual, etc.
    float     confidence;      // 0–1
    uint32_t  ais_mmsi;        // 0 if no AIS match
    char      name[32];        // Vessel name if known
};

// Watch confirmation — every 10 min
struct WatchConfirmMsg {
    uint32_t  seastead_id;
    uint64_t  timestamp_us;
    uint8_t   watch_status;    // Confirmed on-watch
};

Convoy Configuration


// Convoy state — broadcast by leader, 1 Hz
struct ConvoyStateMsg {
    uint32_t  leader_id;
    uint64_t  timestamp_us;
    double    origin_lat;      // Convoy origin position
    double    origin_lon;
    float     course_deg;      // Convoy course
    float     speed_kn;        // Convoy speed
    float     grid_spacing_m;  // Current grid spacing
    uint8_t   num_seasteads;
    struct {
        uint32_t id;
        float    grid_east_m;  // Target position relative to origin
        float    grid_north_m;
        uint8_t  status;       // active, joining, leaving, disabled
    } members[MAX_FLEET_SIZE];
};

10. Cost Summary

10.1 Per-Seastead Hardware Cost

Category	Item	Budget	Recommended	High-End
Communications	Directional WiFi (×4)	$220	$400	$1,000
Communications	Network switch + cabling	$80	$150	$250
	2.4 GHz omni backup antenna	$40	$60	$80
	LoRa 900 MHz emergency	$30	$50	$80
	Communications subtotal	$370	$660	$1,410
Navigation	RTK GPS (ZED-F9P + antenna)	$250	$250	$2,500
	Dual-antenna heading kit	$100	$150	$300
	IMU (9-axis)	$30	$50	$300
	Navigation subtotal	$380	$450	$3,100
Maritime Safety	AIS Class B transponder	$350	$500	$800
Maritime Safety	VHF marine radio (DSC)	$200	$350	$500
	Maritime Safety subtotal	$550	$850	$1,300
Sensors	Marine radar (compact solid-state)	$1,200	$2,000	$4,000
	PTZ cameras (×4) + mounts	$400	$1,200	$3,000
	IR/low-light camera upgrade	$0	$500	$2,000
	Sensors subtotal	$1,600	$3,700	$9,000
Computing	Main computer (Raspberry Pi 5 / Jetson)	$80	$250	$700
	AI vision processor (Jetson Orin)	$0 (use main)	$500	$700
	UPS, power conditioning, enclosure	$100	$200	$400
	Computing subtotal	$180	$950	$1,800
	Starlink terminal	$300	$600	$2,500
	TOTAL per Seastead	$3,380	$7,210	$19,110

10.2 Recurring Costs per Seastead

Item	Monthly Cost	Notes
Starlink service	$120 – $250	Priority / Mobile plan for maritime use
Software updates / cloud services	$10 – $50	Weather routing, chart updates, fleet management cloud
Total recurring	$130 – $300

10.3 Fleet Software Development

Component	Estimated Effort	Cost (contractor)	Notes
Convoy formation manager	2–3 months	$20k – $40k	Grid management, join/leave, leader election
Object tracking & fusion	3–4 months	$30k – $60k	Multi-sensor fusion, parallax ranging, track database
Watch & consensus system	1–2 months	$10k – $25k	Watch manager, Raft consensus, UI dashboard
Mesh network management	1–2 months	$10k – $20k	OSPF config, link monitoring, failover
Autopilot integration	3–6 months	$30k – $80k	Thruster allocation, station keeping, stabilizer control
UI / dashboard	2–3 months	$20k – $40k	Map display, alerts, watch status, configuration
Testing & integration	2–3 months	$20k – $40k	Sea trials, debugging, optimization
Total software development	14–23 months	$140k – $305k	Can be reduced by using open-source components (ArduPilot, ROS, etc.)

💡 Cost Reduction Opportunities

ArduPilot: The open-source ArduPilot autopilot supports boat/rover mode with GPS waypoint navigation, and has support for motor allocation. Building on this could save 3–6 months of autopilot development.
ROS 2: The Robot Operating System provides communication middleware, sensor drivers, and navigation packages that could accelerate development significantly.
Open source tracking: Libraries like SORT, DeepSORT, and OpenCV provide object tracking foundations.
Phased approach: Start with basic formation keeping + comms ($3k/seastead), add AI tracking and watch systems later.

10.4 Fleet Cost Examples

Fleet Size	Hardware (Recommended)	Software (One-Time)	Annual Recurring	Cost per Seastead (Yr 1)
5	$36,050	$200,000	$12,000	$49,210
10	$72,100	$200,000	$24,000	$29,610
20	$144,200	$200,000	$48,000	$19,610
50	$360,500	$250,000	$120,000	$14,610
100	$721,000	$300,000	$240,000	$12,610

Note: Year 1 cost per seastead includes the software development amortized across the fleet, plus hardware and one year of recurring costs. In subsequent years, the cost drops to just hardware (if expanding) plus recurring fees.

11. Recommended Next Steps

Phase 1: Prototype (Months 1–3)

Build networking hardware kit for 2 seasteads
Deploy ZED-F9P RTK GPS with RTKLIB
Implement basic position broadcast over UDP multicast
Test WiFi link range over water at planned spacing
Cost: ~$1,500 for hardware + 1 month engineering time

Phase 2: Two-Seastead Formation (Months 3–6)

Implement basic formation keeping (hold grid position)
Deploy AIS transponders and basic radar
Test walkway connection protocol
Validate mesh routing under real conditions
Cost: ~$5,000 additional + 2 months engineering

Phase 3: Multi-Seastead Convoy (Months 6–12)

Scale to 4–8 seasteads
Implement watch system and consensus protocol
Deploy camera AI for object detection
Implement parallax ranging
Test joining/leaving protocols
Cost: ~$30k–$80k engineering + per-seastead hardware

Phase 4: Full Convoy (Months 12–24)

Scale to full fleet
Implement weather routing and storm avoidance
Fine-tune wave interaction spacing
Full watch rotation system
Community features (inter-seastead video, shared events)
Cost: Remaining software budget + fleet hardware