ARM Gazebo

This package contains the Gazebo simulation used in the RoboCup ARM challenge.

Setup

Note: Skip this section if you downloaded the ARM Gazebo Virtual Machine.

• Install docker engine (not docker Desktop!!!) (tested on v. 19.03, 20.10)

  Usually, this should work

    sudo apt install docker.io
  

  or install from binaries

    https://docs.docker.com/engine/install/binaries/
  

  See also Post-installation steps for Linux. In particular, add your user to the docker group and log out and in again, before proceeding.

    sudo usermod -aG docker $USER
  

• Install docker-compose (tested on v. 1.28)

  First remove any other docker-compose file, if present (check with which docker-compose)

  Download binay file for v. 1.28.5

    cd /usr/local/bin
    sudo wget https://github.com/docker/compose/releases/download/1.28.5/docker-compose-Linux-x86_64
    sudo mv docker-compose-Linux-x86_64 docker-compose
    sudo chmod a+x docker-compose
    docker-compose -v
  

• Nvidia driver

  Install nvidia-docker2

  NVidia Container Toolkit

• Additional Tools

    sudo apt install gawk
  

Build or pull the docker image

Note: Skip this section if you downloaded the ARM Gazebo Virtual Machine.

To get the docker image you can either build it from the Dockerfile or pull it from `dockerhub'.

Option 1: Build docker image

cd docker
./build.bash

Option 2: Pull image

docker pull iocchi/arm_gazebo:latest

Run

cd docker
./run.bash [x11|nvidia|vnc]

x11 default mode uses X11 server, nvidia if you have NVidia graphic card and you installed nvidia-docker, vnc if you have problems with other modes uses a virtual xserver accessible through a web browser at http://localhost:3000/

Note: if you are using the ARM Gazebo Virtual Machine, run in standard mode ./run.bash

Explanation of the startup procedure

The script run.bash use docker-compose to start the docker containers with the configuration given in the files docker/dc_<*>.yml

To see the running docker containers, use the command

docker ps

The container armgazebo is the one that is running the ARM gazebo simulation

The container is running a tmux session. You can attach to this session with the following command

docker exec -it armgazebo tmux a

The container initialize the gazebo simulation by running the script bin/init_sim.bash. In this script you can find the commands sent to the tmux server, in particular the following windows are created: 0) gazebo simulator, 1) moveit, 2) rviz (command prepared but not launched), 3) scripts to spawn objects.

Note: you can create additional tmux windows with CTRL-b c. You can detach from tmux with CTRL-b d

If you need to restart the simulation, move to the first window of tmus (the one running gazebo), quit the simulation and other ROS nodes with 'CTRL-c' in this window, relaunch the simulator and all the ROS nodes (press the UP key to access the previous command and then ENTER).

Configure and spawn objects

Create a config file in config folder with objects described in this format

<object_name> <type> <x> <y> <z> <yaw> <pitch> <roll>

Spawn objects from the docker container

First, enter the docker container

docker exec -it armgazebo tmux a

Then, issue the following commands to manage objects in gazebo simulator

cd ~/src/arm_gazebo/scripts
python gazebo_objects.py -a <object config file>

For example:

python gazebo_objects.py -a ../config/ARM1.txt

You can add more config files in the config folder

To reset a world, you can use the script

./reset_world.bash <object config file>

This script will first delete all the objects according to the patterns in file config/ARMd.txt and then spawn the new objects in the specified config file.

For example, this command will reset the world with the configuration ARM1.txt

./reset_world.bash ../config/ARM1.txt

For other interactions with gazebo objects, see a list of options with

python gazebo_objects.py -h

Options
-h	help
-l	list objects
-m	list available models
-a <obj> <type> <x> <y> <z> <yaw> <pitch> <roll>|<filename>	add an object or all objects in config file
-d <obj>|<filename>|all	   delete one object, all objects in config file, or all objects in the scene
-w	world properties
-s <obj>	object properties and state

Rviz visualization

From the docker container

docker exec -it armgazebo tmux a

you can create a new tmux windows with CTRL-b c and then run

cd ~/src/arm_gazebo/config
rosrun rviz rviz -d robot_camera.rviz

You can detach from tmux with CTRL-b d

Troubleshooting

• In case of errors due to existing containers, use this command to clean them

    docker container prune -f