eDO_automatic_controller

This package is an upper layer allowing to easily control the robot using Moveit.

Installing using docker (simulation only)

The catkin repository will be created in the container, so do not clone this folder inside your catkin repo

Before doing anything, you must add execution permission to the xserver to run GUI application in docker In your terminal run
xhost +

In the .docker folder, execute the command line
docker-compose build

Then execute
docker-compose up

The docker instance is up and a name <the_container_name> has been given to it. You should see
Attaching to <the_container_name> in your terminal.
For instance
Attaching to docker_edo_gazebo_1 in your terminal.

now you can access open a new terminal inside the container using
docker exec -it <the_container_name> /bin/bash
Let's suppose <the_container_name> is docker_edo_gazebo_1. Then the command to enter is
docker exec -it docker_edo_gazebo_1 /bin/bash

Open 2 terminals using
docker exec -it <the_container_name> /bin/bash

In the first terminal run the make_workspace.sh script. This file will just build the catkin repository using catkin build and add to your .bashrc what you need If this does not work, make sure you set the permissions to run the script
./edo_automatic_simulator/make_workspace.sh

run the command
roslaunch edo_automatic_simulation init_simulation.launch This will launch the gazebo simulation and moveit server made by Stefan Profanter

In the second, you can launch your own script to control the robot. For instance, enter the command
rosrun edo_automatic_simulation move_down.py

Summary : Terminal 1
xhost +
docker-compose build
docker-compose up
Terminal 2
docker exec -it <the_container_name> /bin/bash
./edo_automatic_simulator/make_workspace.sh
roslaunch edo_automatic_simulation init_simulation.launch
Terminal 3
rosrun edo_automatic_simulation move_down.py

Installing on your workstation

ROS dependencies

This package does not need more dependencies than the packages edo_control or edo_gazebo

For the previous packages, at least Moveit is necessary.

Python dependencies

This package uses the python packages :

• numpy
• numpy-quaternion

Simulation

Clone the directory into your catkin directory. Your workspace should contain the following packages

repository	repository url
edo_gripper	https://github.com/Pro/edo_gripper
edo_gripper_moveit	https://github.com/Pro/edo_gripper_moveit
edo_gazebo	https://github.com/Pro/edo_gazebo
edo_description	https://github.com/Pro/eDO_description

Real robot

Clone the directory into your catkin directory. Your workspace should contain the following packages

repository	repository url
edo_control/edo_control_v3	https://github.com/ymollard/eDO_control or https://github.com/Bracewind/eDO_control_v3
edo_moveit	https://github.com/Pro/eDO_moveit
edo_core_msgs	https://github.com/Comau/eDO_core_msgs

To connect to the real robot, use the description given in the edo_control repository

How to use it ?

Simulation

• launch the file init_simulation.launch
• wait that everything has started
• launch your code

Real robot

• Calibrate your eDO using the application
• Configure your terminal (see this file for more details)
• launch the file init_connexion.launch
• wait that everything has started
• launch your code

How to write your own code ?

The code written should only use the functions given by the EdoAbstractClass.

• The class EdoDummy is a stub used to test programs using the EdoAbstractClass
• The class MoveitSimulator is used for controlling the robot with the simulator
• The class MoveitController is used for controlling the real robot

Code examples on how to write a simple program using the controllers are given in the example folder