Space is Crowded: Building a Real-Time Satellite Collision Predictor

With thousands of satellites and pieces of debris orbiting Earth, "space traffic control" is no longer science fiction—it’s a data engineering challenge. For my latest project, I decided to bridge the gap between astrodynamics and web development by building the On-Orbit Collision Predictor.

The Problem

Monitoring satellite conjunctions (close approaches) requires processing Two-Line Element (TLE) sets from NORAD, propagating their orbits, and calculating the probability of a "handshake" at 17,500 mph.

My Tech Stack

• The Math: I used MATLAB with SGP4/SDP4 propagation models for the heavy-duty orbital mechanics and risk assessment.

• The Backend: A Python/Django RESTful API to automate the ingestion and parsing of live NORAD data feeds.

• The Frontend: A Next.js dashboard featuring interactive 3D visualizations to make the math understandable at a glance.

Key Takeaways

1. Automation is King: By automating the TLE ingestion, I moved from manual data entry to a real-time monitoring system.

2. Visualization Matters: Using 3D graphs to represent collision probability transformed raw coordinates into actionable insights.

3. Cross-Disciplinary Engineering: Combining specialized tools like MATLAB with modern web frameworks like Next.js showed me how to build bridges between scientific research and user-facing software.

Check it out on my GitHub: https://github.com/raymondoyondi/On-Orbit-Collision-Predictor