/ardrone_gazebo

This ardrone_gazebo repository contains a ROS Noetic package to simulate the old but well known AR.Drone.

Primary LanguageC++

The ardrone_gazebo ROS Noetic package

This ardrone_gazebo repository contains a ROS Noetic package to simulate the old but well known AR.Drone. The ROS package was created by incorporating ideas, code and artefacts from tum_simulator, tahsinkose/sjtu_drone and NovoG93/sjtu-drone combined with a (partial) clean-up and rewrite of a few source code files.

A single AR.Drone in an empty Gazebo world:

Features

  • Simulates the AR.Drone in a standard ROS Noetic setup with Gazebo 11.
  • Tries to mimic the original tum_simulator. (Currently far from perfect!)
  • Supports multiple concurrent AR.Drones in Gazebo!
  • All required external Git repositories are located on the PXLRoboticsLab GitHub community.

Dependencies

How to use

ROSNoeticDocker container

This ardrone_gazebo package, and all its dependencies, are included in the ROSNoeticDocker repository. That repository was created for the AI & Robotics courses at the PXL University of Applied Sciences and Arts. But it's free for all to use!

It's the most convenient way to get up and running with the ardrone_gazebo package. Just clone the ROSNoeticDocker repository, build and run the container and you're ready to start the AR.Drone in Gazebo.

Manual install in own workspace

If you want to make changes to the ardrone_gazebo package, it's advised to install it in your own ROS Noetic workspace. (Don't use the ROENoeticDocker to do this, unless you're comfortable purging the ardrone_gazebo build statements form the Dockerfile before building the image.)

Prerequisites

  • Basic bash, git & ROS knowledge
  • An installed and working ROS Noetic distribution (ros-noetic-desktop is recommended)
  • Gazebo 11 installed
  • A sourced Catkin workspace

Installation steps

Setting up the ardrone_gazebo package in your own workspace is super easy. Just clone two repositories, build and source the workspace.

<in your workspace dir> $ cd src
<in your workspace dir>/src $ git clone https://github.com/PXLRoboticsLab/ardrone_autonomy
<in your workspace dir>/src $ git clone https://github.com/PXLRoboticsLab/ardrone_gazebo
<in your workspace dir>/src $ cd ..
<in your workspace dir> $ catkin_make
<in your workspace dir> $ source devel/setup.bash

After this, you're ready to start the AR.Drone in Gazebo.

Running the simulator

There are a few launch files included with this repository.

A single drone

<On your ROS system>$ roslaunch ardrone_gazebo single_ardrone.launch

Two drones

<On your ROS system>$ roslaunch ardrone_gazebo two_ardrones.launch

Four drones

<On your ROS system>$ roslaunch ardrone_gazebo four_ardrones.launch

Ten drones

<On your ROS system>$ roslaunch ardrone_gazebo ten_ardrones.launch

Context:

Each academic year there's an one-week international Drone course at the PXL University of Applied Sciences and Arts (Belgium). Unfortunately, during the previous few years it was canceled due to the well-known COVID-19 pandemic. Luckily, (it looks like) the saga has come to an end!

In the past, we (I'm one of the researchers/lecturers responsible for the course) used the AR.Drone. The students are a broad mix of computer science students. Some have a technical focus others are more business oriented. The existing Gazebo plugins for the AR.Drone are outdated or I couldn't make them work on Noetic. Using an old ROS Distribution and Python2 in 2022 was a big no-go! Moving to the PX4 wasn't an option due to the type of students. Therefore, I mixed and matched code, artifacts, ... from a few repositories, cleaned-up/rewrote some code... And, made the ardrone_gazebo package. It's nothing special. Big thank you to the TUM Computer Vision Group, tahsinkose and NovoG93! They did the real work.

  • Because multi-agent-systems is one of the topics I teach, I took this opportunity to add multi-drone support.

Enjoy!