Starting up

Please follow these instructions when switching on Care-O-bot 3. This page describes how to physically start the robot and what software needs to be launched. Moreover, it provides checklists to ensure a successful startup.

Before Checklist

Ensure the following

  • The robot is fully charged

  • The robot’s wireless emergency stop is fully charged

  • The robot’s tablet is fully charged

Hardware

  1. Remove the charger from the robot (red plug)

  2. Turn the key on the robots back side to the right, hold it for three seconds, and then release it

    robot powers on and its computers boot up

  3. Wait for bootup to complete

    ubuntu login-sound can be heard

  4. Release the two emergency buttons at the robot (rotate and pull)

  5. Release the wireless emergency button

    1. Release red cap on wireless emergency device

    2. Press and hold green button on wireless emergency device

      green light starts to blink

    3. Turn power key on robot clockwise and release it after a few seconds

      you hear clicking noises

    4. Release green button on wireless emergency device

    5. Repeat the last three steps

      you hear “emergency button released” (only if the ROS components on the robot’s first computer are already running, see below)

Software

Some components (i.e., basics, navigation, perception) have to be started on different PCs of the robot, others (e.g., visualization, behaviour control) on your desired study/demo computer.

First robot computer

The following components all have to run on cob3-5-pc1 to enable the robot’s basic functionalities and navigation (optionally):

Note

You need to log into the robot for each of the following commands individually using the following in a new terminal window:

$ ssh cob3-5-pc1

  1. Start ROS communication

    $ roscore

  2. Start the bringup component to initialize the robot

    $ roslaunch cob_bringup robot.launch

  3. Start the script server to execute commands/scripts on the robot

    $ roslaunch cob_script_server script_server.launch

  4. Start the navigation component (if you want the robot to drive)

    $ roslaunch cob_navigation_global 2dnav_ros_dwa.launch

Second robot computer

The following components have to run on cob3-5-pc2 to enable robot vision:

  1. Log into the robot’s second computer

    $ ssh cob3-5-pc2

  2. Start the image republisher (if you want camera images from the robot)

    $ roslaunch accompany_siena_images_republisher accompany_images_republisher.launch

Study/demo computer

You can use any computer that is connected to the house’s network and correctly configured to use the robot’s roscore at http://cob3-5-pc1:11311 (i.e. running Ubuntu 12.04 precise and ROS hydro) to start behavioral components. In most cases though, you would probably prefer to use the legacy workstation inside the robot house office, named rh-desktop-n as the demo user as this should be configured correctly, has RViz and the dashboard installed, and allows you to access the robot via Secure shell without having to type in a password.

  1. Start RViz for a robot visualization

    $ roslaunch cob_bringup rviz.launch

  2. Start the dashboard for command execution (e.g., initialization, postures)

    $ roslaunch cob_bringup dashboard.launch

  3. Initialize robot modules you intend to use in the dashboard’s Command GUI

  4. Start any additional components that you are using for behavior generation

After Checklist

Ensure that RViz is in sync with the robot

  • Tray, arm and body are displayed in the correct position (after their initialization)

  • The robot’s position on the map is correct (if you use the navigation)