Obtaining the Code

Clone the code repository to your home directory using Git:

cd ~
git clone https://gitee.com/your-repo/drone_autonomy.git

Alternatively, download the compressed workspace package and extract it directly to your home directory.

Workspace Compilation

Move the drone_autonomy_ws directory from the cloned repository to your home directory, then compile the ROS workspace:

cd ~/drone_autonomy_ws
catkin_make
source devel/setup.bash

System Context

This codebase enables autonomous navigation for drones using the PX4 flight stack, ROS MoveBase framework, Intel T265 visual-inertial camera, and LD06 single-line lidar. It supports both real hardware deployment and PX4 simulation environments.

Sensor Configuraton

1. T265 Setup: Configure the Intel T265 to publish odometry data to MAVROS. Refer to official ROS-PX4 integration tutorials for calibration and topic mapping.
2. Lidar Integration: Launch your LD06 lidar driver and confirm the LaserScan topic is active. Update the costmap configuration in drone_autonomy_ws/src/drone_nav/config/costmap_common_params.yaml to match your lidar's coordinate frame and topic:

   observation_sources: lidar_scan
   lidar_scan: {sensor_frame: lidar_link, data_type: LaserScan, topic: /ld06/scan, marking: true, clearing: true}
   

Launch File Details

• cartographer_slam.launch: Starts Cartographer SLAM backend (headless, no RViz visualization)
• mapping_launch.launch: Launches Cartographer with RViz for real-time 2D mapping
• move_base_core.launch: Loads core MoveBase navigation parameters (internal use only)
• mapping_nav.launch: Enables simultaneous mapping and autonomous navigation
• local_nav_only.launch: Local navigation mode without pre-built maps (uses real-time sensor data for obstacle avoidance)
• simulation_launch.launch: Integration with PX4 Software-In-The-Loop (SITL) simulation

Key Source Files

• px4_movebase_bridge.cpp: Core node that converts MoveBase's velocity commands into MAVROS-compatible messages for PX4. It includes conditional logic to handle real hardware and simulation environments, with detailed comments in the code.
• waypoint_navigator.cpp: Implements waypoint-based autonomous navigation. The default setup targets a single waypoint; extend the code by adding addditional waypoints to the predefined list for multi-point missions.

Operational Workflow (Local Navigation Example)

1. Ensure MAVROS is connected to the PX4 flight controller and T265 odometry is active.
2. Launch the local navigation stack:

   cd ~/drone_autonomy_ws
   source devel/setup.bash
   roslaunch drone_nav local_nav_only.launch
   

3. Open a new terminal to start the PX4-MoveBase bridge node:

   cd ~/drone_autonomy_ws
   source devel/setup.bash
   rosrun drone_nav px4_movebase_bridge
   

4. The drone will automatically take off and hover at approximately 0.6m altitude. Use RViz's 2D Nav Goal tool to send navigation targets.

Extension Options

• Alternative Localization: Replace the T265 with lidar-based SLAM (e.g., GMapping) combined with barometric altitude. Publish the odometry pose to mavros/vision_pose/pose for integration with PX4's EKF.
• Visual-Inertial Odometry: Integrate VINS-Fusion by publishing its pose estimates to mavros/vision_pose/pose for use in PX4 navigation.