SlayterHIL
A hardware-in-the-loop flight controller testbed for Embedded Systems @ Purdue.
SlayterHIL is a hardware-in-the-loop (HIL) platform that lets us run a drone flight controller on real hardware while the “world” around it (sensors + physics + environment) is simulated.
The payoff is safe, repeatable, automated testing for embedded flight code, without crashing a real drone.
Why this matters
Flight controllers are a perfect embedded “stress test”:
- Real-time constraints (timing matters)
- Multiple subsystems at once (IMU + comms + control loops + motor outputs)
- Safety-critical behavior (you want strong validation before real flight)
SlayterHIL is also meant to feel like industry:
- small, well-scoped tickets
- CI-driven regression testing
- clear interfaces between teams
- reusable infrastructure for future club projects
The big idea (in one picture, conceptually)
Test Automation defines a scenario → Flight Simulation generates realistic sensor outputs → Test Node delivers them deterministically to the DUT (flight controller) → DUT outputs motor commands → simulation updates the world → automation decides pass/fail.
What we’re building
Core components
- DUT (Device Under Test): the flight controller firmware being validated
- Test Node: real-time bridge between sim and DUT (deterministic injection / fast I/O)
- Flight Simulation: physics + sensor generation (turn motor outputs into sensor streams)
- Test Automation: defines tests, runs them locally/CI, and evaluates pass/fail
Hardware & software stack
Hardware
- Raspberry Pi (server-side orchestration; model TBD)
- ESP32-S3 (test node / interface layer)
- Sensors planned for validation + contracts:
- LSM6DSO IMU
- MPL3115A2 barometer/altimeter
Software
- C / C++
- Zephyr RTOS (on the flashing/interface device/test node components)
- PyTest (test definitions + assertions)
- CI/CD via GitHub Actions (chosen over Jenkins for accessibility + ease of use)
What’s been accomplished (Fall 2025)
By end of Fall 2025, the team:
- finalized the overall system design
- onboarded tooling + workflows
- eliminated major blockers and clarified responsibilities
- built early test artifacts and interface prototypes
