keanucz

keanucz / detour

Public

On-board AI agents autonomously saving satellites from orbital debris @ treehacks 2026

141
14
100% credibility
Found Feb 17, 2026 at 129 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
TypeScript
AI Summary

Detour is an interactive dashboard that visualizes satellite orbits, detects nearby space debris threats, and uses AI agents with real physics to recommend collision avoidance maneuvers.

How It Works

1
🛰️ Discover Detour

You hear about Detour, a friendly helper that watches for space junk threatening your satellite and suggests safe paths around it.

2
🚀 Open the dashboard

Click to launch the colorful 3D globe showing Earth and your satellite's path in real time.

3
📡 Pick your satellite

Enter your satellite's ID or draw a custom orbit to start tracking its journey.

4
⚠️ Spot nearby dangers

Red dots pop up showing risky debris closing in, with countdowns to closest passes that grab your attention.

5
🤖 Call for AI help

Tap the button to let smart helpers scan threats and dream up clever dodges.

6
🛡️ Get safe maneuver plans

Watch as the AI suggests gentle pushes to steer clear, checking fuel and safety first.

Orbit secured

Your satellite glides safely past the junk, with a clear report confirming no more worries.

Sign up to see the full architecture

5 more

Sign Up Free

Star Growth

See how this repo grew from 129 to 141 stars Sign Up Free
Repurpose This Repo

Repurpose is a Pro feature

Generate ready-to-use prompts for X threads, LinkedIn posts, blog posts, YouTube scripts, and more -- with full repo context baked in.

Unlock Repurpose
AI-Generated Review

What is detour?

Detour runs on-board AI agents that autonomously scan for orbital debris threats to satellites, compute collision risks using real physics like CW equations and covariance propagation, then plan delta-v maneuvers while checking fuel and constraints. You interact via a TypeScript Next.js dashboard with 3D globe views of orbits, debris swarms, and threat trajectories, plus a live terminal for natural-language queries like "scan ISS for 24h threats." Built for edge hardware like NVIDIA GX10, it delivers operator briefs in seconds, demoing full autonomy from scout to execution.

Why is it gaining traction?

The multi-agent pipeline—scanning catalogs, assessing Pc probabilities, optimizing along-track burns, safety-checking—feels production-like, with SSE streaming for real-time terminal updates. Paired with manual orbit tweaks and sim controls, it hooks devs wanting instant "what if" debris avoidance without spinning up custom sims. 122 stars from Treehacks 2026 buzz show appeal for agents handling 2026-era space traffic.

Who should use this?

Satellite ops engineers prototyping autonomous detours from LEO debris, AI builders testing LangGraph on physics tools, or space startups mocking kanban boards on GitHub for threat triage. Perfect for recast detour GitHub forks into custom scrum boards on GitHub or dev detour workflows.

Verdict

Impressive demo for agent-driven orbital safety, but 1.0% credibility and 122 stars flag hackathon maturity—expect tweaks for real TLE/CDM feeds. Fork for space sims now; watch for polish.

(198 words)

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.