How themaxnet/hiveparafoil works
Project type and purpose
• This is an embedded‐software project for an Arduino/ESP32‐based GPS navigation system (parafoil guidance).
• The firmware reads NMEA sentences from a GPS module, applies a Kalman filter for smoothing/estimation, and likely outputs processed position data (e.g., on telemetry or SD card).
• The accompanying directories (“Docs” with filter specs and NMEA manual, “Misc” with test scripts, “Oakleigh GPS Test” data files) support development, calibration, and post‐processing.
File structure
Top‐level directories:
• Docs – contains Kalman filter and NMEA reference manuals (design inputs).
• Misc – Python test script (testnav.py) for verifying navigation logic against recorded data.
• Oakleigh GPS Test – raw CSV, MATLAB, and text files for real‐world GPS data.
• Old – archived versions of ESP32/Arduino sketches demonstrating earlier iterations.
• parafoilarduino – the current, active Arduino sketch (parafoilarduino.ino).
• readme.md – points to the current firmware location.
Build/deployment specifics:
• The .ino file compiles via the Arduino or PlatformIO toolchain, targets an ESP32/Arduino board.
• No explicit CI or Dockerfiles are present, so deployment is local to the development machine.
README for additional insights
• README is minimal: directs to the main Arduino script.
Main components & their relationships
a. Hardware Layer
• GPS sensor (UART output of NMEA sentences)
• Microcontroller board (Arduino/ESP32)
• Optional peripherals: SD card module (for logging), telemetry radio or USB serial
b. Firmware Layer (on Arduino)
• UART driver – reads raw NMEA strings
• NMEA parser – tokenizes and extracts GGA/RMC fields
• Kalman filter module – applies state estimation (from Docs/Kalman filter.pdf)
• Data logger / Telemetry output – writes filtered data to SD or streams over serial
• Configuration/constants – filter tuning parameters, system timing
c. Off‐board Tools
• Python test script (Misc/testnav.py) – ingests logged or sample data, verifies filter accuracy
• MATLAB/CSV data sets (Oakleigh GPS Test) – used by the test script to benchmark performance
Interaction flow:
GPS sensor → Serial input → NMEA parser → Kalman filter → {Data logger, Telemetry}
Logged data + Oakleigh test sets → Python script → Performance plots
Architectural patterns & principles
• Layered processing pipeline: sensor interface → parsing → estimation → output
• Single MCU monolith – no service‐oriented or distributed microservices, but clear modular separation in code (parser module, filter module, comms module).
• Reusability: archived “Old” sketches demonstrate iterative design and fallback points.
Detail
• Every key processing block in the Arduino firmware.
• Configuration sources (Kalman docs, NMEA manual).
• Test harness outside the MCU boundary.
• Don’t overthink – more granular separation (parser vs. filter vs. logger) yields clarity and easier future extensions.