youBot_from_scratch

This repository offers a detailed description to perform autonomous navigation using a KUKA youBot controlled via an external Laptop. This work was performed at B-it-bots@Work Lab in Hochschule Bonn-Rhein-Sieg.

Motivation

The intention behind performing this work is to get in-depth knowledge of How to configure a robot with ROS to perform autonomous navigation from scratch. For this purpose, we have used a KUKA youBot and an external Laptop.

Pre-requisites

• A KUKA youBot

• A laptop with the following specifications:

  • Ubuntu v18.04
  • ROS Melodic Morenia

Steps to follow

Create a catkin workspace

• First create a workspace, we call it youBot_from_scratch and then build it.

mkdir youBot_from_scratch
cd youbot_from_scratch
mkdir src
catkin build
cd src

• Inside src, we now need to clone the following packages:

  • youbot_driver: git clone https://github.com/youbot/youbot_driver.git

  • youbot_description: git clone https://github.com/youbot/youbot_description.git

  • youbot_driver_ros_interface: git clone https://github.com/youbot/youbot_driver_ros_interface.git

  • youbot_applications: git clone https://github.com/youbot/youbot_applications.git

Configuring Hardware

Note: Power the youBot and make sure to turn-off the internal PC.

KUKA youBot drivers

• We'll install the youBot drivers and test-run the youBot:

  • Build the youbot_driver package:

  cd youbot_driver
  mkdir build
  cd build
  cmake ..
  make
  

  • We need to edit some configuration files in youbot_driver package:

    • Connect the ethernet cable from KUKA youBot to your Laptop

    • Check your PC's Ethernet device ID using ifconfig (It is usually something like eth0).

    • Go to config folder inside youbot_driver package, and open youbot-ethercat.cfg in edit-mode and changeEthernetDevice = <Your Ethernet Device ID>

  • Navigate to the youbot_application package and build it using:

  cd youbot_applications
  mkdir build
  cd build
  cmake ..
  make
  

  • Export the respective package paths so that it can be discovered by other packages in ROS:

  echo "export ROS_PACKAGE_PATH=\$ROS_PACKAGE_PATH:/home/<user>/youBot_from_scratch/src/youbot_driver:\
  /home/<user>/youBot_from_scratch/src/youbot_applications" >> ~/.bashrc
  

  Note: Change the tag to your default system user and change the workspace name if different.

  • Now we'll check if the drivers are installed successfully by running a demo script.

    cd youbot_applications/hello_world_demo/bin
    sudo ./<name_of_binary_file>
    

Hokuyo lidar sensor

• We'll integrate the Hokuyo Lidar Sensor with our external laptop:

  Note: Since we are using an external laptop instead of KUKA youBot's internal PC, we need to connect the Hokuyo Lidar sensors to the USB port of the laptop.

  • Connect the sensor to the Laptop and check if the sensor is found by typing dmesg in a terminal window. Keep a note of its port ID (something like ttyACM0)

  • Check if you have the rights to open the port:

  od /dev/<port ID>
  

  • If it doesn't work, try it with super user permission:

  sudo  od /dev/<port ID>
  

  • if it works with super user permission, then add yourself to the dialout group:

  sudo adduser <YOUR_USERNAME> dialout
  

  • Log out and log back in. Check if you're in the dialout group:

  groups
  

  • Get the device ID/Sensor ID:

  rosrun urg_node getID /dev/ttyACM0 --
  
  

  • Write udev rules to give Hokuyo consistent device names:

    cd /etc/udev/rules.d
    

    • Create a following rule file:

    sudo vim 80-youBot.rules
    

    • Paste the following rule and save it:

    SUBSYSTEMS=="usb", KERNEL=="ttyACM[0-9]*", ACTION=="add", ATTRS{idVendor}=="15d1", ATTRS{idProduct}=="0000", MODE="666",      PROGRAM="/opt/ros/melodic/lib/urg_node/getID /dev/%k q", SYMLINK+="sensors/hokuyo_%c", GROUP="dialout"
    

Running youBot with ROS

• Create a catkin package inside the src :

  catkin_create_pkg youBot std_msgs roscpp rospy
  

Create youBot driver node

• In order to work with youBot, we first need to fire up the youBot drivers:

  • Create a launch file youBot_driver.launch in the launch directory of the youBot package and paste the following code:

  <?xml version="1.0" encoding="UTF-8"?>
  <!launch>
    <include file="$(find youbot_driver_ros_interface)/launch/youbot_driver.launch" />
  </launch>
  

  Note: The youBot driver launch file is already provided with the youbot_driver_ros_interface package. Here, we are simply calling that launch script.

Create youBot Hokuyo sensor node

• To visualize the output of Hokuyo Lidar sensor:

  • Create a launch file youBot_sensor.launch in the launch directory of the youBot package and paste the following code:

  <?xml version="1.0" encoding="UTF-8"?>
  <launch>
    <!-- Front Laser scanner-->
    <node name="hokuyo_node_1" pkg="urg_node" type="urg_node" respawn="false" output="screen">
    <param name="port" type="string" value="/dev/sensors/hokuyo_H1102594"/>
    <param name="intensity" type="bool" value="false"/>
    <param name="min_ang" value="-2.356194437"/>
    <param name="max_ang" value="2.35619443"/>
    <param name="cluster" value="1"/>
    <param name="frame_id" value="base_laser_front_link"/>
    </node>
    <!-- Back Laser scanner-->	
    <node name="hokuyo_node_2" pkg="urg_node" type="urg_node" respawn="false" output="screen">
    <param name="port" type="string" value="/dev/sensors/hokuyo_H1838912"/>
    <param name="intensity" type="bool" value="false"/>
    <param name="min_ang" value="-2.356194437"/>
    <param name="max_ang" value="2.35619443"/>
    <param name="cluster" value="1"/>
    </node> 
    <!-- Launching RViz -->
    <node name="rviz" pkg="rviz" type="rviz"/>
  </launch>
  

  Note: 1. The value of the parameter port needs to be change according to your sensor ID (you can find it by navigating to /dev/sensors/hokuyo_xxxxxxx), 2. Depending upon whether you have 1 or 2 hokuyo sensors, you can edit the launch file.

Create youBot urdf node

• To import the youBot in RViz,

  • Create a launch file youBot_urdf.launch in the launch directory of the youBot_package and paste the following code:

  <?xml version="1.0" encoding="UTF-8"?>
  <launch>
      <!-- Launching urdf -->
      <param name="robot_description" command="$(find xacro)/xacro --inorder '$(find youbot_description)/robots/youbot.urdf.xacro'" />    
      <!-- Start TF publishers -->		  
      <node pkg="robot_state_publisher" type="robot_state_publisher" name="robot_base_state_publisher" output="screen" />
  </launch>
  

Create youBot teleop node

• In order to move the youBot around in the workspace, we need to launch teleop package. We'll be using a Joystick controller to move the robot around.

  • Create a launch file youBot_teleop.launch in the launch directory of the youBot_package and paste the following code:

  <?xml version="1.0" encoding="UTF-8"?>
  <launch>
      <node pkg="joy" type="joy_node" name="joystick" output="screen"/>
      <node pkg="teleop_twist_joy" type="teleop_node" name="teleop" output="screen">
      <param name="enable_button" value="5"/>
      </node>
  </launch>
  

  Note: The parameter enable_button allows to enable regular-speed movement. We set its value as 5 because it is mapped with the L1 Key in Joystick and we use it as a standard key to enable regular-speed movement in all of our KUKA youBot (We have plenty of them!).

Create SLAM gmapping node

Note: In order to do Simultaneous Localization And Mapping, we need to launch the gmapping package in ROS.

• Create a launch file youBot_mapping.launch in the launch directory of the youBot_package and paste the following code:

<?xml version="1.0" encoding="UTF-8"?>
<launch>
  <node name="slam_gmapping" pkg="gmapping" type="slam_gmapping">
    <remap from="/scan" to="/scan"/>
    <param name="base_frame" value="base_footprint"/>
  </node>
</launch>

Mapping 'the Arena'

• Build the catkin workspace:

catkin build

• To start mapping, we need to launch the following launch-files:

  `roslaunch youBot youBot_drivers.launch`
  `roslaunch youBot youBot_urdf.launch`
  `roslaunch youBot youBot_sensor.launch`
  `roslaunch youBot youBot_teleop.launch`
  `roslaunch youBot youBot_mapping.launch`

• Move the youBot around cautiously using the Joystick controller and create a map of the whole arena.

• After you are done with mapping, you can simply save the map by doing Ctrl+C in the terminal where you have launched youBot_mapping node. The map files will be saved in your package.

• It's a good practise, to create a map directory inside the youBot package and move both the newly generated files there <map>.yaml and <map>.pgm.

Creating AMCL Node

Note: In order to do localization, we need to launch the amcl package in ROS.

• Create a launch file youBot_localization.launch inside the launch directory of the youBot package and paste the following code:

  <launch>
    <!-- MAP Server Node-->
    <arg name="map_file" default="$(find youbot)/map/test_env.yaml"/>
    <node name="map_server" pkg="map_server" type="map_server" args="$(arg map_file)" />
    <!-- AMCL Node -->
    <node name="amcl" pkg="amcl" type="amcl" output="screen">
        <remap from="scan" to="/scan"/> <!--according to package-->
        <param name="odom_frame_id" value="odom"/>
        <param name="odom_model_type" value="omni"/>
        <param name="base_frame_id" value="base_footprint"/> <!--according to package-->
        <param name="global_frame_id" value="map"/>
        <!-- If you choose to define initial pose here -->
        <param name="initial_pose_x" value="0"/>
        <param name="initial_pose_y" value="0"/>
        <param name="initial_pose_a" value="0"/>
    </node> 
  </launch>
  

• To start localization, we need to launch the following launch-files:

   `roslaunch youBot youBot_drivers.launch`
   `roslaunch youBot youBot_urdf.launch`
   `roslaunch youBot youBot_sensor.launch`
   `roslaunch youBot youBot_teleop.launch`
   `roslaunch youBot youBot_localization.launch`
  

  • You can move the youBot around and visualize the localization of the youBot in the RViz window by subscribing to appropriate topics.

  ---

  Todo: Relevant images, rviz config file, navigation