// OVERVIEW
About this project
An autonomous-drone flight-simulation environment built on PX4 SITL and Gazebo. It stands up the full software-in-the-loop stack — a PX4 flight controller, a physics-accurate quadcopter, and simulated sensors — then layers autonomous mission planning, waypoint navigation, and live telemetry on top, so flight behavior can be developed, flown, and tested end to end in simulation before any of it touches real hardware.
// ARCHITECTURE
System architecture
Python mission scripts talk to a PX4 SITL flight controller over MAVLink (via MAVSDK). PX4 flies a quadcopter model inside Gazebo, which simulates the airframe's physics and feeds back simulated GPS / IMU / barometer data. The estimator and controller run the real PX4 firmware, so behavior in the sim maps directly onto behavior on hardware. Telemetry streams back out for logging and monitoring.
// FEATURES
Key features
Software-in-the-Loop Stack
Runs PX4 SITL against a Gazebo-simulated airframe, so the exact flight-controller firmware flies a physics model of the drone — the same code path that runs on real hardware.
Autonomous Missions
Programmatic mission planning and waypoint navigation over MAVLink: arm, take off, fly a route, and land, all driven from Python instead of a manual controller.
Simulated Sensors
GPS, IMU, and barometer feed the estimator exactly as they would in flight, so navigation and control logic is validated against realistic sensor behavior.
Live Telemetry
Streams position, attitude, battery, and flight-mode telemetry out of the sim in real time for logging, monitoring, and debugging each run.
Repeatable Test Environment
The whole stack spins up from scripts, so scenarios are reproducible — the same mission replays to compare behavior as the control logic changes.