Skip to content
SP
Colour theme
All projects

Intelligent systems · Academic project

Academic project

Autonomous Drone Simulation

PX4 SITL, ROS 2, and Gazebo for autonomous control and telemetry validation.

Conceptual illustration created for this case study — not a screenshot of the application.

Overview and context

An academic intelligent-systems project simulating autonomous drone control with PX4 SITL, ROS 2, MAVROS, MAVLink, and Gazebo, focused on control workflows, sensor validation, and diagnosing failures from logs.

Problem

Autonomous flight logic needs a safe environment to expose state-estimation and preflight reliability problems before hardware is involved.

My role and contribution

  • Developed and validated the autonomous drone-control simulation.
  • Implemented real-time telemetry and OFFBOARD control workflows.
  • Validated IMU, GPS, and positioning data.
  • Performed system-level debugging and log analysis to investigate state-estimation and preflight reliability issues.

Technical approach

  • PX4 SITL with Gazebo provides the flight dynamics and sensor simulation.
  • ROS 2 and MAVROS bridge control commands and telemetry over MAVLink.
  • OFFBOARD control workflows drive autonomous behaviour during simulated flight.
  • Flight logs are the primary diagnostic surface for state-estimation and preflight failures.

Evidence and outcomes

  • A working simulation stack exercising autonomous control end to end.
  • Identified and investigated state-estimation and preflight reliability issues through log analysis.