Note, if you feel like not using this path but installing directly on your system, you are free to do it, but you are on your own.
Here are the instructions to have the simulation system running.
-
Download VREP from here, V-REP PRO EDU V3.5.0 rev4
- extract folder anywhere, where you preferer
- make sure you have the the version 3.5, other versions may not work properly
-
Download
Robobo_Scene.ttt
andRobobo_Model.ttm
files and place them inside your VREP folder- place
Robobo_Scene.ttt
file inVREP/scenes
- place
Robobo_Model.ttm
file inVREP/models/robots/mobile
- place
-
Download docker
-
Download git + Editor
-
Download repository
git clone https://github.com/ci-group/learning_machines_robobo.git
Now you downloaded everything, you can start the environment
-
Start docker
- Linux/systemd: start docker daemon
- OsX/Windows: open docker app
-
Start docker container (this will open a terminal with python and all required libraries installed). This step could take a while, since first your computer will download the docker image from the internet, secondly will compile some stuff from your project folder.
- Linux/OsX, open a terminal in the learning_machines_robobo project folder and type:
./start-docker.sh
- Windows, open CMD and change directory to the learning_machines_robobo project folder, and type:
start-docker.bat
- Linux/OsX, open a terminal in the learning_machines_robobo project folder and type:
-
Start VREP
- open the program
- open the scene file that you downloaded before: Robobo_Scene.ttt
- enable Real Time button (optional)
- start the simulation with the play button
-
Run the script
- inside the docker container running, type
./src/send_commands.py
-
Change the script, experiment and HAVE FUN :D
If you are having problems with Windows HOME, follow this guide it may fix your issues
You need to change the file start-docker.bat
changing from
SET project_folder=%~dp0
docker run --rm -it -v %project_folder%:/root/projects --net=host cigroup/learning-machines bash
to
SET project_folder=/c/Users/YourUser/path/to/learning_machines_robobo
docker run --rm -it -v %project_folder%:/root/projects --net=host cigroup/learning-machines bash
where your project is in the folder C:\Users\YourUser\path\to\learning_machines_robobo
. Change this variable accordingly to your own username and path to the project folder. You cannot use folders outside C:\Users\
and remember to change all backslasesh (\
) to forward slashes (/
).
- start docker with
docker run --rm -it -v "${PROJECT_FOLDER}:/root/projects/" --net=host cigroup/learning-machines bash
- change folder to
/root/projects/
(inside docker) - run
catkin_make install
(only run for the first setup, it's not needed even across different docker containers) - run
source devel/setup.bash
- run
src/send_commands.py
command to test the connection
The environment variable ROS_MASTER_URI
idendifies where the ROS MASTER NODE is running. Verify that your phone is in "Local ROS Master" and adjust the IP of the ROS_MASTER_URI
variable with the address of the phone (visible at the first page of the application).
The library we provide is already taking care of this enviroment variable, but in case you encouter any problems you know where to look for.