GPS Satellites Tracker
v0.1.57 — Real-Time GPS Tracking Station
GPS SATELLITES<br>TRACKER
A real-time GPS satellite tracking dashboard using data from USB GPS Receivers<br>Docker containers and a lot of coffee and cake
Live Dashboard
Explore Features
Parquet Log Files Created
GB Total Data So Far
GPS Satellites Seen
Days to Build
Cups of Earl Grey
Cakes Consumed
Key Features
Real-Time Satellite Intelligence
A comprehensive GPS tracking dashboard with 3D globe visualisation, signal analysis, and real world satellite telemetry.
3D Globe View
Interactive Three.js globe displaying satellite positions, orbital paths, and real-time tracking with multiple camera perspectives, zoom and pan controls
Signal Analysis
Real-time SNR monitoring with bar charts, line plots, skyplot radar view, and historical sparkline telemetry for each satellite.
Satellite Directory
Browse all GPS satellites with spacecraft specifications, launch history, constellation block details, and real-time status.
Historical Playback
Time-travel through logged GPS data with a timeline slider, variable playback speeds, and frame-by-frame analysis.
Receiver Management
Track multiple GPS receiver locations simultaneously with coordinate display, error scatter plots, and CEP accuracy metrics.
Real-Time Telemetry
Live SSE streaming of GPS data with auto-refresh, instantaneous satellite position updates, and signal quality monitoring.
System Architecture
How It Works
A five-container architecture: USB serial data ingestion via MQTT, visualised through a Flask web frontend.
USB GPS Receiver
Cheap USB GPS module connected via serial port, outputting standard NMEA 0183 over serial.
Backend Pipeline
Raspberry Pi parsing NMEA data, computing satellite positions, and publishing structured telemetry to MQTT brokers.
Processing Service
Docker based Python processing service producing structured parquet files from the MQTT subscriptions
Web Dashboard
Modern Flask application serving Three.js 3D globe, DuckDB-powered Parquet queries, and SSE streaming for live updates.
Coming Soon
Build or buy a tracking device to stream your own telemetry and geo-location data directly to the server.
Future Hardware Support
Support for the battery powered LILYGO TTGO T-A7670G ESP32 based board with global 4G LTE and GPS.
Get Started
Try the Live Dashboard
Explore the full GPS tracking interface with real-time satellite data, 3D globe visualisation, and signal analysis tools.
Launch Live Dashboard
About
The Project
A weekend hobby project turned experiment in agentic coding — building a personal GPS tracking station from scratch.
Creator
Robert Wolf
Machine Learning Engineer
YouTube
[email protected]
System Information
Created Date:<br>8th June 2026
Software Version:<br>v0.1.57
Parquet Log Files:<br>46 files
Total Data Size:<br>7.24 MB
Date Coverage:<br>2026-06-09 to 2026-06-19
Receiver Location:<br>53.383°N, 2.173°W
The Story
Inspired by a recent Veritasium video on GPS jamming over Europe and Canada "Something is jamming GPS over Europe. Here's what we found", I decided to kick off a weekend hobby project to build a personal GPS tracking station. My goal wasn't just to log data, but to run an experiment in using agentic coding to help me.
I picked up a cheap USB GPS receiver, fired up my development server running TrueNAS SCALE, and used Google's Antigravity CLI to build a two-container architecture: a backend to ingest USB serial data, and a friendly web frontend. Within three hours and about 50 to 60 micro-revisions through the CLI, I had a working Docker container logging GPS data into hourly Parquet files. I never touched a single line of backend code in that first three house and the code quality was good.
Next up was the WebUI. I built it panel by panel, feature by feature completely through the Antigravity CLI. While the AI nailed vague prompts perfectly most of the time, it also went down some bizarre rabbit holes. I frequently had to step in as the "Senior Dev" to course-correct.
Out of the entire build, only a couple of bugs regarding the time slider defeated the agent, requiring me to step in and manually tweak the code.
I will be updating the website from time to time. If you have any problems please wait a few minutes as I'm probably updating the code and docker container behind the scenes.
Technology Stack
Python 3.13<br>Flask<br>Gunicorn<br>Three.js<br>DuckDB<br>Parquet<br>MQTT<br>Docker<br>TrueNAS SCALE<br>pynmea2<br>pyserial<br>Antigravity CLI<br>OpenCode<br>SSE<br>Orbitron UI<br>paho-mqtt<br>pyarrow
GPS-Tracker
Tracking Dashboard
A full-featured GPS satellite monitoring interface with three interconnected views: 3D globe, signal noise analysis, and satellite directory.
TACTICAL 3D
Orbital perspective with live satellite positions, orbital paths, and real-time tracking overlays
ERROR HEAT MAP
GPS accuracy error distribution with colour-coded heat overlay showing positional uncertainty zones
EARTH VIEW
Clean Earth perspective...