Every eVTOL startup I talk to has the same problem: flight test data scattered across spreadsheets, CSVs, and custom scripts that break every other week. No standard format. No validation. No easy way to query what happened during Test 47.
I hit this wall myself. So I built aero-telemetry, a lightweight, open-source telemetry ingestion API for aerospace prototyping.
What It Does
- Ingest sensor data via HTTP: battery state, GPS, IMU, motor RPM
- Validate everything: voltage ranges, motor counts, coordinate bounds
- Query by session, time range, aircraft
- Batch ingest for time-series analysis
- Zero config: SQLite default, runs on a laptop
Why FastAPI + Pydantic
FastAPI gives async request handling and auto-generated OpenAPI docs. Pydantic gives type-safe validation with clear error messages. Together they mean I spend zero time on boilerplate and all time on aerospace logic.
Example: battery voltage below 10V triggers a validation error. For eVTOL packs running 400-800V, a 3.7V reading is either a sensor fault or a wrong unit and the API catches it before it corrupts your dataset.
Architecture
Three core models:
- Aircraft: static config (mass, propulsion type, MTOM)
- TestSession: a single test event (hover, transition, endurance)
- TelemetryPoint: sensor snapshot with optional nested data
SQLite storage with SQLAlchemy 2.0 async. Repository pattern for clean separation. 26 pytest tests covering CRUD, validation, and edge cases.
Demo
2-minute walkthrough:
https://www.loom.com/share/58a0bf51b5794c879bed101aaac82753
Current Status
v0.1.0 released. Apache-2.0 licensed.
GitHub: github.com/Mohammed-Fayaz-Ahamed/aero-telemetry
What's Next
- PostgreSQL support for production deployments
- WebSocket streaming for real-time telemetry
- Grafana dashboard integration
- Time-series aggregation endpoints
Who This Is For
eVTOL startups doing prototype testing. Drone fleet operators. Graduate researchers building UAM simulations. Anyone who needs structured flight test data without enterprise overhead.
If you're building in this space, I'd love your feedback what's missing, what's wrong, what would you actually use?
Top comments (1)
Great breakdown. An API gateway is one of those components that looks simple from the outside but becomes a critical control plane as systems grow. Routing, authentication, rate limiting, observability, and request transformation all need to work together without becoming a bottleneck.
I like the focus on the filter chain pattern because it provides a clean way to add cross-cutting concerns while keeping business services focused on their core responsibilities.
The biggest challenges in production usually come from edge cases: failure handling, latency overhead, security policies, and maintaining consistent behavior across services. Great architectural walkthrough — these fundamentals are valuable for anyone designing distributed systems.