Building a semi-submersible platform with differential thrust requires a blend of marine navigation, robotics, and industrial monitoring software. Below are the recommended open-source solutions.
PyPilot is indeed the best starting point for a custom build. It is designed to run on low-power hardware like the Raspberry Pi and is highly modular.
servo.py or pilot scripts to translate the heading error into two distinct PWM signals (one for each motor) rather than a single rudder angle.If you find PyPilot too limited for multi-motor mixing, ArduPilot is the industry standard for autonomous vehicles. The "Skid-Steer" (tank drive) configuration in ArduRover is mathematically identical to differential thrust for a boat.
Frigate is a locally-hosted Network Video Recorder (NVR) that uses AI (via Google Coral or CPU) to detect people and objects. This is much more efficient than traditional motion detection, which triggers every time a wave splashes.
If you prefer a more traditional CCTV interface, Shinobi is modern and written in Node.js, making it very responsive in a browser. ZoneMinder is the veteran of the space, highly capable but with a steeper learning curve.
This is the standard open-source chart plotter. It will provide the visual GUI for your position, AIS (if you have a receiver), and weather GRIB files. It can communicate with PyPilot via the SignalK protocol.
Think of this as the "Central Nervous System." SignalK takes data from your solar chargers (NMEA 2000 or Modbus), your motors, and your GPS, and translates it into a single data format that all your other software can read.
While often used for houses, it is perfect for seasteads. Use it to monitor: