Autonomous Drone Control and Localization Project - WIP

Project Overview

This project is designed to implement and test drone control and localization functionalities using ROS 2 and PX4 using a Raspberry Pi 4 companion computer. This includes path following, AR tag detection, and position estimation.

Directory Structure

vanguard_ws/
├── src/
│   ├── control/            # Control package
│   │   ├── control/
│   │   │   ├── __init__.py
│   │   │   ├── path_following.py
│   │   │   ├── hover.py
│   │   │   ├── arm_disarm.py
│   │   │   ├── translation_control.py
│   │   ├── setup.py
│   ├── localization/       # Localization package
│   │   ├── localization/
│   │   │   ├── __init__.py
│   │   │   ├── ar_tag_detection.py
│   │   │   ├── position_estimation.py
│   │   │   ├── config/
│   │   │   │   ├── ar_tag_positions.yaml
│   │   ├── setup.py
│   ├── px4_msgs/           # PX4 messages package
│   │   ├── px4_msgs/
│   │   ├── setup.py

Dependencies

• ROS 2 (Humble)
• PX4 Autopilot
• Python 3.8+
• Required ROS 2 packages: px4_msgs, sensor_msgs, geometry_msgs
• OpenCV
• YAML
• ROS2 V4L2 Camera Node
• DepthAI ROS Driver

Installation Instructions

1. Install ROS 2 and PX4:

   • Follow the ROS 2 Humble installation guide.
   • Install PX4 following the PX4 Autopilot installation guide.

2. Clone the Repository:

   git clone <repository_url> vanguard_ws
   cd vanguard_ws

3. Install Dependencies:

   sudo apt update
   sudo apt install python3-opencv python3-yaml

4. Build the Workspace:

   colcon build

5. Source the Workspace:

   source install/setup.bash

Usage

Path Following

1. Configure Path:

   • Edit the path.yaml file in src/control/control/ to define paths to follow.

2. Run Path Following Node:

   ros2 run control path_following

AR Tag Detection

0. Define AR Tag Locations for Localization:

   • Edit the ar_tag_positions.yaml file in src/localization/localization/config/ to define AR tag positions.

1. Run AR Tag Detection Node:

   ros2 run localization ar_tag_detection

Position Estimation

1. Run Position Estimation Node:

   ros2 run localization position_estimation

Scripts Description

Control Package

• path_following.py:

  • Implements the path following functionality using trajectory setpoints.
  • Reads path waypoints from a YAML file and follows the defined path.

• arm_disarm.py:

  • Implements arming, disarming, takeoff and landing of the drone.
  • Publishes appropriate commands to arm and disarm the drone.

• trajectory_control.py:

  • Implements trajectory control functionalities.
  • Allows setting custom trajectories for the drone to follow.

• hover.py:

  • Implements hover functionality.
  • Maintains the drone's position at a specified point.

Localization Package

• ar_tag_detection.py:

  • Detects AR tags in images and publishes the relative position of the drone to the detected tags.
  • Uses OpenCV for AR tag detection.

• position_estimation.py:

  • Estimates the global position of the drone based on AR tag detections and odometry data.
  • Uses AR tag positions defined in a YAML file for accurate localization.

Configuration Files

• ar_tag_positions.yaml:
  • Contains the positions of AR tags used for localization.
  • Define each tag's position in the global frame.

Authors

• Immanuel Ampomah Mensah
• Contributors: Sarah McMahon, Joshua Huang, Smruti Suresh, Noreen Hussein, Sakina Saidi

License

This project is licensed under the MIT License.