<![CDATA[Challenges in medical logistics, such as geographical limitations, urban congestion, and ground transport risks, necessitate efficient and safe delivery solutions. Unmanned Aerial Vehicles (UAVs) offer a promising alternative. This research aims to develop and implement an autonomous UAV navigation system capable of complex multi-landing missions, validated for medical delivery scenarios. The system employs a hierarchical control architecture, integrating the Robot Operating System 2 (ROS2) as the high-level mission supervisor with the PX4 flight controller for low-level control. Mission logic, including navigation using the Waypoint Follower algorithm and landing sequences, is formally governed by a Finite State Machine (FSM) implemented in ROS2. Communication between ROS2 and PX4 is facilitated by Micro XRCE-DDS. System validation was performed via Software-in-the-Loop (SITL) simulation using Gazebo, featuring a mission scenario with four waypoints and one intermediate landing. Simulation results demonstrated that the system successfully completed the entire mission autonomously (100% success rate). Quantitative analysis of position accuracy revealed total position errors () between 1.35 m and 1.57 m during navigation phases, and an error of 0.84 m at the intermediate landing. Errors were found to be consistently dominant on the Y-axis. In conclusion, the proposed architecture, which separates mission logic (FSM in ROS2) from flight control (PX4), provides a robust and functional approach for autonomous UAV applications requiring complex mission interactions.]]>
