A browser-based spatial intelligence and surveillance simulator.

This prototype explores the concept of “sousveillance” and spatial data fusion by layering real-time, open-source intelligence (OSINT) feeds over a 3D navigable globe. By combining live telemetry with military-grade WebGL shaders, it transforms public data into an intelligence dashboard.

We are moving beyond language models into spatial intelligence—systems that understand physical relationships, continuous movement, and change over time. This project is a foundational data engine that ingests raw telemetry (flights, satellites, cameras) and makes the physical world queryable in real-time.

• Real-Time Data Fusion Engine: A FastAPI backend that acts as a secure proxy, ingesting and parsing complex public data streams to avoid frontend CORS issues and rate limits.
• Live Orbital Mechanics: Pulls raw Two-Line Element (TLE) sets from CelesTrak and calculates live satellite coordinates using the sgp4 physical modeling library.
• ADS-B Flight Tracking: Real-time aircraft positioning and callsign tracking using OpenSky Network telemetry.
• Tactical WebGL Shaders: Custom Cesium.PostProcessStage fragment shaders that hijack the rendering pipeline to compute screen-space luminance, mapping it to:
  • Night Vision: High-contrast phosphor green with CRT scanlines and animated film grain.
  • FLIR Thermal: Heat-signature simulation utilizing a localized color ramp and high-luminance blooming.
• Spatial Video Projection: Drapes live looping video feeds onto specific geographical coordinate geometry (Test Zone: Hunter St & Bethune St intersection).
• “God Mode” HUD: A responsive, absolute-positioned tactical overlay featuring live mouse coordinate targeting and sensor array toggles.

Backend (Data Engine):

• Python 3
• FastAPI & Uvicorn (Lightweight asynchronous web server)
• sgp4 (Earth satellite orbital tracking)
• requests (API ingestion)

Frontend (Render Engine):

• CesiumJS (3D Geospatial Globe)
• WebGL (Custom Fragment Shaders)
• Vanilla JavaScript, HTML, CSS (Zero-dependency UI)

To run this intelligence platform locally on your machine:

1. Clone the repository:

   git clone https://github.com/Forest-code-ai/Space.git
   cd Space/vibe

2. Create/activate the virtual environment & install dependencies (PowerShell):

   ./scripts/bootstrap.ps1
   . ./.venv/Scripts/Activate.ps1
   pip install -r requirements.txt

3. Spin up the data engine and web server:

   python -m vibe

4. Initialize the uplink:

   Open your browser and navigate to http://localhost:8000.

Built as a proof-of-concept for geospatial data pipelines and multi-agent AI development.

If you want to push README updates to your portfolio:

git add README.md
git commit -m "docs: added comprehensive README outlining spatial architecture"
git push