MRS UAV Gazebo Simulation

Metapackage for the MRS UAV Gazebo simulation pipeline.

Installation

1. Install the MRS UAV System
2. sudo apt install ros-noetic-mrs-uav-gazebo-simulation

Submodules

Package	Repository
MRS Gazebo Common Resources	mrs_gazebo_common_resources
PX4	px4_firmware
PX4 SITL Gazebo	px4_sitl_gazebo

Unmanned Aerial Vehicles

Model	Spawn argument	Simulation
DJI f330	--f330
DJI f450	--f450
Holybro x500	--x500
DJI f550	--f550
Tarot t650	--t650
T-Drones m690	--m690
NAKI II	--naki

Starting the simulation

Use one of the prepared Tmuxinator sessions:

• tmux/one_drone_gps

Customizing drone using the "spawner parameters"

TODO the wiki page https://ctu-mrs.github.io/docs/simulation/drone_spawner.html contains potentially duplicit or contradicting information. These two pages should be merged and the information should be presented at only one of the locations.

The UAV platforms can be additionaly equipped with sensors (rangefinders, 2d lidars, stereo cameras etc.).

Start of the Gazebo simulator

To start the prepared example of Gazebo world call:

roslaunch mrs_simulation simulation.launch world_file:='$(find mrs_gazebo_common)/worlds/grass_plane.world' gui:=true

At this point the Gazebo world will only contain the environment with grass plane but with no vehicles yet.

Spawning of UAVs

The simulation.launch will automatically start the mrs_drone_spawner python node. If you use a custom launch file to start the simulation, you can start it separately:

roslaunch mrs_simulation mrs_drone_spawner.launch

The mrs_drone_spawner will perform the following tasks:

• Spawn vehicle models in the Gazebo simulation (ids from 0 to 250). This is done internally by calling the command rosrun gazebo_ros spawn_model.

• For each vehicle, PX4 firmware and mavros is started at specific port numbers depending on the vehicle ID.

Vehicles are added to the simulation by calling the spawn service of the mrs_drone_spawner. The service takes one string argument, which specifies the vehicle ID, type and sensor configuration. Example: spawn a single vehicle with a down-facing laser rangefinder:

rosservice call /mrs_drone_spawner/spawn "1 --enable-rangefinder"

To display the manual containing a list of all available arguments, perform a dry-run of the script:

rosrun mrs_simulation mrs_drone_spawner

The arguments are also listed in the mrs_simulation/config/spawner_params.yaml file. Note that not all sensors are available for all the vehicle types. The config file stores the available configurations in the following format: parameter: [default_value, help_description, [compatible_vehicles]]

Multiple vehicles may be spawned with one service call:

rosservice call /mrs_drone_spawner/spawn "1 2 3 4 5 --t650 --enable-bluefox-camera --enable-rangefinder"

Spawn position may be specified by a command line argument --pos x y z heading: [m, m, m, rad]

rosservice call /mrs_drone_spawner/spawn "1 --f550 --enable-rangefinder --pos 10 -15 0.3 0.7"

For multiple vehicles, --pos defines the spawn point of the first vehicle. Following vehicles will be spawned in a line, with an x-offset of 2 meters from the previous vehicle.

Spawn position may be specified by a .csv or a .yaml file using --file absolute_path_to_file:

rosservice call /mrs_drone_spawner/spawn "1 --f450 --enable-rangefinder --enable-ouster --use-gpu-ray --ouster-model OS1-64 --file `pwd`/spawn_poses.yaml"

Use a whitespace instead of an ID to get an available ID automatically assigned by the spawner.

rosservice call /mrs_drone_spawner/spawn " --f450 --enable-rangefinder"