This tool is for research, policy analysis, and education only. It is not operational SSA, a collision-warning system, a maneuver-command tool, or flight-safety decision support.
AI-LEO Policy Simulator: Quick Tutorial
Local dashboard walkthrough for the Phase-C prototype.
1. What the simulator does
It builds synthetic low-Earth-orbit populations and propagates them with SGP4. It screens geometric close-approach candidates, applies three proxy governance regimes (R0, R1, R2), and supports bounded AI triage for analyst prioritization. A local Next.js dashboard reads exported results from a local FastAPI service. Outputs are for research evidence and education, not flight operations.
2. Choose a track
Track 1 — Understand without installing
- Read the 10-second workflow on the landing page.
- Compare R0/R1/R2 governance regimes (short summary below; full text on the landing page).
- View screenshots (reviewed local-run images).
- Download quickstart notes and public-facing manifests.
- Review evidence-pack notes and the planned DOI citation route.
Track 2 — Run locally as a developer
The public website is a static explainer. The full simulator (FastAPI backend + dashboard) runs locally until public-backend security is completed.
3. Static demo vs local simulator
Static demo
- Runs entirely in the browser.
- Uses pre-generated 1000-satellite sample data.
- Safe for the public website.
- Good for quick understanding and screenshots.
Local simulator
- Runs the Python backend and Next.js dashboard.
- Can regenerate scenarios.
- Useful for researchers and thesis work.
- Available after public repo cleanup.
4. R0 / R1 / R2 in one minute
These regimes are simplified policy abstractions for research comparison, not real legal rules or operational maneuver procedures.
- R0 — uncoordinated baseline (no shared coordination rule).
- R1 — soft coordination (assign responsibility mainly to the more capable actor).
- R2 — fairness-aware enforcement (consider capability and accumulated burden).
Full definitions and the mini-table are on the landing page under “Governance regimes compared.”
5. Two-minute local run (developer track)
Developer quick start:
- Clone the public repository.
- Install Python and Node dependencies.
- Start the local backend.
- Start the dashboard.
- Open the dashboard and run the 100-satellite demo.
Terminal 1:
git clone <PUBLIC_REPOSITORY_URL>
cd ai-leo-policy-simulator
python -m uvicorn phaseC.backend.main:app --host 127.0.0.1 --port 8000Terminal 2:
cd ai-leo-policy-simulator/phaseC/frontend
npm install
npm run devOpen: http://localhost:3000
6. One-click workflow inside the dashboard
- Select scenario
visual_analytics_demo_100sat(default). - Click Run 100-sat demo.
- Wait for dashboard outputs to refresh.
- Open
/orbits,/policy,/ai-triage, and/evidencefrom the sidebar.
7. What each page means
Dashboard: summary metrics and the run button.
Orbits: SGP4-propagated sample visualization.
Conjunctions: research-grade close-approach screening candidates (geometric screening, not operational alerts).
Policy: R0 / R1 / R2 proxy governance comparison on the same candidate set.
AI Triage: simulated unresolved-action prioritization for analyst review workflows.
Evidence: figures, JSON exports, and reproducibility pointers documented in the repository.
8. What not to do
- Do not interpret outputs as operational conjunction alerts.
- Do not use for flight operations or real-time safety decisions.
- Do not expose the FastAPI backend publicly without a security review.
- Do not treat AI output as maneuver or command advice.
9. Troubleshooting
-
Backend offline: start
python -m uvicorn phaseC.backend.main:app --host 127.0.0.1 --port 8000from the repo root. - No data: run the visual analytics pipeline or use Run 100-sat demo so exports exist for the scenario.
-
Frontend build issue: run
npm installthennpm run buildinphaseC/frontend. -
Python dependency issue:
pip install -r requirements.txtfrom the repo root.