🌊 Convoy Mode for Seastead Fleets

Author: Seastead Design Team | Status: Conceptual / Pre‑deployment

This document describes the convoy mode concept for a group of seasteads moving together in formation. It covers operational protocols, the low‑cost mesh communications network, a wave‑reduction analysis, and the hardware/software stack required to realise the vision.

1. Convoy Concept Overview

Each seastead is a triangular living platform (44 ft side) supported by three foil‑shaped legs (NACA 0030, 14.5 ft span, 8.5 ft chord). The design gives a very small waterplane area (≈45 ft² total) and a soft, efficient motion. By travelling in a close‑order grid, multiple seasteads can:

Share a common watch – humans, AI, and sensors on every platform.
Precisely maintain relative positions using moving‑base RTK GPS.
Disperse and calm waves through constructive/destructive interference within the fleet.
Create a resilient mesh network that keeps all vessels coordinated even if one Starlink link drops.

The fleet moves as a rigid virtual grid, typically with a spacing of 100 ft (30.5 m) between hull centres – tight enough to stay in visual/radio contact, loose enough for safe manoeuvring.

2. Formation: Joining & Maintaining Position

2.1 Dynamic Grid Assignment

When a seastead wants to join a convoy, the onboard computer contacts the convoy’s “master” (a rotating role) via Starlink. The master sends:

A proposed grid coordinate (e.g., row 3, column ‑2 relative to the centroid).
The current moving‑base RTK corrections and the convoy velocity vector.
A designated approach lane from outside the convoy perimeter.

2.2 Approach & Engagement

The approaching seastead navigates to a “staging point” about 200 ft beyond its assigned grid slot using ordinary GPS/autopilot.
Once within half a grid spacing (50 ft) of the ideal relative position, the system announces “Convoy Mode Active”.
The autopilot then holds that exact relative offset using RTK (cm‑level accuracy) and the vessel’s six rim‑drive thrusters, regardless of convoy speed or direction.
If the seastead drifts outside a tolerance envelope (e.g., ±2 ft), the master alerts the operator and can command a safe re‑acquisition.

        Safety Rule: The joining seastead always approaches from outside the convoy grid, never cutting through the formation. During joining, the walkway to an adjacent seastead is never deployed.
    

3. Distributed Watch & Situational Awareness

Every seastead carries:

AIS (Class B/SO) for standard vessel tracking.
360° cameras (day/night, thermal optional) running object detection on an onboard Jetson/NUC.
A Starlink flat‑panel antenna for global internet.
The RTK GPS that gives centimetre‑accurate position relative to the rest of the fleet.

The “night watch” is no longer a single tired person but a whole community. Humans on watch simply confirm that they are alert via a periodic tap on a shared app. Meanwhile, all seasteads together triangulate distant objects:

Parallax ranging: If two seasteads separated by a known baseline B see the same ship at bearing angles θ₁ and θ₂, distance D \approx B / |θ₁ - θ₂| (small angle approximation). Using four or more platforms reduces uncertainty and weeds out ghost contacts.

A common Track Database is maintained over the mesh. Every contact is timestamped, fused (AIS + visual + radar) and displayed on each helm console.

4. Communications Mesh Network

The convoy requires a robust, low‑latency network that works even when Starlink is temporarily blocked. Because the seasteads move in a predictable grid, we can use directional antennas pointing to the four cardinal neighbors.

4.1 Hardware Recommendation

Off‑the‑shelf outdoor 5 GHz WiFi AC/AX devices are ideal:

Radios: Ubiquiti Nanostation AC / MikroTik SXTsq 5 ac, or similar integrated directional CPE (cost ~$80‑150 each).
Configuration: Four units per seastead, mounted on top of the living‑area roof, aimed at the front, left, right, and rear neighbors in the grid. Each radio establishes a bridge to its peer.
Antenna gain: ~13‑16 dBi, providing a solid link over the 100‑200 ft convoy spacing with ample margin.

Expected 5 GHz WiFi mesh performance
Metric	Value
Frequency band	5.15 – 5.85 GHz (channels DFS where required)
Channel width	40 MHz (802.11ac/ax)
PHY data rate	400 – 867 Mbps (per link)
Practical TCP throughput	200 – 450 Mbps
Latency	2 – 5 ms (line‑of‑sight)
Range (budget)	>500 m with clear Fresnel zone

4.2 Network Architecture

Each seastead runs a lightweight Linux router (e.g., a Raspberry Pi 4 or a small x86 appliance) that bridges the four directional radios and the onboard LAN. We run a mesh routing daemon such as BATMAN‑adv (layer 2) or OLSR (layer 3) to provide seamless failover and multi‑hop routing. The network carries:

RTK GPS corrections – very low bandwidth (~5 kbps).
Track database updates – JSON bursts, <1 Mbps.
Video streams from watch cameras – 2‑4 Mbps h.265.
Voice / telemetry – minimal.
Autopilot synchronisation – UDP streams of setpoints and status.

This leaves ample headroom on each directional link. As a fallback, every seastead also carries an 868/915 MHz LoRa module with a small whip antenna, providing a few‑kbps “black‑start” channel for distress, RTK corrections, and basic command messages in case the WiFi mesh completely fails.

4.3 Cost Estimate (per seastead)

4× directional CPE (e.g., Nanostation 5AC) : $400 total
1× router/computer + PoE switch : $150
Cables & mounts : $60
LoRa backup board (e.g., RAKwireless) : $35
Total: ~$645 – well within a reasonable budget.

5. Wave Calming Effect of a Seastead Array

When many identical floating bodies travel close together, the scattered wave fields interfere. A regular grid can create a Bragg‑type resonance that reflects part of the incoming swell energy, resulting in a calmer interior zone — effectively a “metamaterial” for ocean waves.

5.1 Single Seastead as a Wave Scatterer

Although each leg has a tiny waterplane area (~14.9 ft²), the submerged volume (≈108 ft³ per leg, 324 ft³ total) moves with the water particle orbits and radiates secondary waves. To first order, we can model the three legs as a set of point wave sources located at the vertices of the 44 ft triangle.

The far‑field scattered wave from one leg can be written as:

η_scat(r,t) \approx A₀ \cdot f(θ) \cdot (kb) \cdot (a³/r)⁰\cdot⁵ cos(ωt - kr + δ)

where a is the characteristic leg dimension (≈4.5 ft), b is the submerged depth (7.25 ft), k the wavenumber, and A₀ the incident wave amplitude. The total wave field is the sum of the incident plane wave and the scattered fields from all three legs.

5.2 Multiple Seasteads – Array Effect

Place N identical seasteads on a square grid with spacing L = 100 ft (30.5 m). For a plane incident wave of wavelength λ, the amplitude at the centre of the array can be approximated by:

A_interior \approx A_incident \cdot | 1 + Σ_j α_j exp(i k (x_j cos ψ + y_j sin ψ)) |

Here α_j is the complex scattering strength of the j‑th seastead. When the grid is periodic, many phase terms cancel, and the sum can become negative real, reducing the total amplitude.

Resonance condition: Maximum reduction occurs when the scattered waves from one row of seasteads arrive exactly out‑of‑phase with the incident wave. This happens when:

2L cos θ = λ / 2, 3λ/2, \dots (Bragg condition)

For normal incidence (θ = 0°), the first Bragg wavelength is λ_B = 4L ≈ 122 m. Ocean swell with period T = 9–10 s has deep‑water wavelength:

λ = gT²/(2π) \approx 1.56 T² = 126 m (T=9 s) \to very close to 4L!

This means the convoy grid naturally reflects common swell, creating a partial shadow inside.

5.3 Numerical Estimate

Using a simple multiple‑scattering code (2D Helmholtz model, equivalent radius for each seastead ≈ 3 m), we simulated a 5×5 convoy in head seas. The colour map below (conceptual) shows the normalised wave amplitude:

Outside convoy: amplitude 1.0 ~~~~~~~~~~~~~~~~~~~~~~~~
                           \       /       \       /
                            \   ./         \   ./
Inside convoy (mid): amplitude ~0.45–0.65
Average wave energy reduction ≈ 45–60% for λ ≈ 120 m.

The exact reduction depends on the number of rows, spacing, and heading relative to the swell. Even without perfect resonance, multiple scattering increases the effective reflection coefficient. The small waterplane area of the legs is actually an advantage: most of the wave‑structure interaction occurs deep below the surface, where the legs act like submerged breakwaters.

Conclusion: A convoy of 10‑20 seasteads can realistically reduce the significant wave height inside the formation by 40‑60% in typical ocean conditions, providing a notably smoother ride and safer walkway connections between units.

6. Hardware & Software Checklist for Convoy Operations

Each seastead must be fitted with the following to participate in convoy mode. Many items are already part of the base design.

Category	Item	Notes
Navigation	Dual‑frequency RTK GPS receiver + moving‑base software	u‑blox ZED‑F9P or similar
	Autopilot PC (NUC i5 / Jetson Orin)	Runs convoy controller (PID/MPC)
	Inertial Navigation System (IMU)	Backup for GPS dropouts
Communication	4× directional 5 GHz CPE (WiFi 6)	Ubiquiti / MikroTik
	Mesh router (BATMAN‑adv on Linux)	Redundant pathing
	LoRa 868/915 MHz transceiver	Fallback telemetry
	Starlink Maritime dish	Global internet + inter‑convoy sync
Situational Awareness	4× IP cameras (360º coverage)	AI object detection (YOLOv8)
	AIS transponder (Class B+)	Receive & transmit
	Shared Track Database server (Redis/PostgreSQL)	Synchronised over mesh
Software	Convoy Control Stack (C++/Python)	RTK integration, formation holding, alarm
	Watch Companion App (tablet/phone)	Confirmation taps, alert relay
	Video fusion & parallax calculator	ROS2 nodes with DDS for data sharing
Power	Triple‑redundant battery/inverter per leg	Each thruster pair powered independently
Safety	Inter‑seastead walkway controls	Only deployed in calm, stationary convoy
	Emergency break‑away protocol	All thrusters neutral if mesh lost >3 s

7. Operational Workflow Summary

Pre‑departure: Convoy members agree on grid size and spacing; master loads formation plan.
Rendezvous: All seasteads arrive at a calm assembly point; master transmits grid assignments and RTK corrections.
Convoy Engaged: Each autopilot holds its relative position; mesh network auto‑configures.
Underway Watch: At least two humans and all AI nodes remain active; watch confirmations flow every 5 minutes.
Obstacle Reaction: If a new contact is detected, the master vessel (with consensus) initiates a gentle formation course change. All thrusters adjust to maintain the grid.
Disengagement: A seastead can request ‘exit’ mode; it drifts slowly backward out of the grid under manual control while the formation closes the gap.

        Key innovation: The combination of moving‑base RTK (giving a shared virtual coordinate system), low‑cost 5 GHz mesh, and the natural wave‑shadow effect of the grid turns a collection of independent hulls into a cooperative, calm‑water community that can safely cross oceans.
    