↑top

Connecting a Workstation and Starting the Robot

Your workstation may be a Ubuntu Linux system of your own, or you can use our preconfigured virtual machine.

This page discusses connection of your laptop or workstation or the usage of a VirtualBox to control the robot. If you do not have a linux computer available you can still use a virtual Linux machine to connect to the Magni robot.

Additional information may be found here.

Use of your own Linux machine with WiFi

If you have your own Linux laptop you can use it to connect to the hotspot on the Magni computer using standard Linux interfaces to connect to a network.

Skip ahead to the section 'Connecting to the Robot and Logging In' for laptop.

Using our out-of-the-box virtual machine workstation

You can setup a Linux VirtualBox to communicate with the Magni robot. In this case we have a Linux image that is preloaded with a great many ROS tools so this is a nice way to get going fast if you have the resources to be able to run VirtualBox environment.

In a later section we will explain how to attach the robot to an existing WiFi network and how ROS can take advantage of that to control the robot.

Connecting to The Magni Robot’s network

At this point you should have either your own linux physical machine or the VirtualBox setup described above and we are ready to connect to the Magni WiFi hotspot.

If you have received a Magni with the Raspberry Pi already installed, or loaded the default Raspberry Pi 3 image from downloads.ubiquityrobotics.com, the robot will boot up in WiFi Access Point mode. This is a WiFi mode which provides its own network to which you can connect your workstation. The SSID (network name) is ubiquityrobotXXXX where XXXX is a number letter combination. You should find it under network list; the password (sometimes called the security key) to connect is robotseverywhere. If you can’t find it under available networks, try restarting the robot.

Connecting a Virtual Machine

If you are running under VirtualBox, you will have installed this virtual machine with a bridged network. Thus, the VM will see whatever network your host system is connected to. If your workstation is running, shut it down. Then connect your host system to the ubiquityrobotXXXX network. Now start the workstation (that is, the Ubuntu system running under VBox); it will be connected to the robot’s network ubiquityrobotXXXX. The password (sometimes called the security key) to connect is robotseverywhere.

Connecting to the Robot and Logging In

Now that you are on the robot’s network, you can connect to the robot itself. On your workstation, start a terminal window (Linux shortcut: ctrl-alt-t). In that window, log in by typing

ssh ubuntu@ubiquityrobot.local

The password initially is set to “ubuntu”.

If you have trouble using ubiquityrobot.local, use the IP address, 10.42.0.1 instead. In some environments, this works better.

You may see:

The authenticity of host ‘10.42.0.1 (10.42.0.1)’ can’t be established.
ECDSA key fingerprint is SHA256:sDDeGZzL8FPY3kMmvhwjPC9wH+mGsAxJL/dNXpoYnsc.
Are you sure you want to continue connecting (yes/no)?

Answer yes

Failed to add the host to the list of known hosts (/somepath/.ssh/known_hosts).

Ignore that. You will be asked for the password, which initially is “ubuntu”.

ubuntu@10.42.0.1's password:ubuntu

Now you are connected and logged in.

Welcome to Ubuntu 16.04.3 LTS (GNU/Linux 4.4.38-v7+ armv7l)

Running ROS apps not requiring a Display

Many of the robot applications we supply as demos or some you may be writting yourself do not require a laptop if they do not need to display anything. These types of command line applications can be run on a console to the robot.

A common simple app to control robot movement for example:

rosrun teleop_twist_keyboard teleop_twist_keyboard.py

Even some complex applications such as to start the robot’s navigation software can be run on the robot directly in a text console.

roslaunch magni_demos simple_navigation.launch

We would suggest that to run simple_navigation you should see our Fiducial-Based Localization Page as our intent here is not to explain that complex mode, simply to show an example of a ROS app.

Setup Laptop ROS Environment For Magni Control

In order to have your laptop be able to issue ROS messages to control the Magni such as running twist on the laptop or to run programs that require a graphical display such as rviz you will need to run ROS on your laptop and setup the laptop instance of ROS so that the robot is the ROS master.

Once the laptop (or VM) is configured with the robot as the ROS master you can inspect the ROS topics on the laptop and even view what the camera is seeing from your laptop or do other things like configure dynamic ROS parameters which all are done with a laptop display normally.

Rather than duplicate the required setup here you should refer to the ROS Workstation Setup Page to do these required configuration steps.

At this point you will be able to control the robot from the workstation keyboard or by using Robot Commander.

Doing a clean Shutdown Of the Magni Robot

Forgive us for first explaining how to shutdown and then power off the Magni robot but we felt this step may be missed if it were at the very end of this page.

We recommend doing clean shutdown on a Magni to lower your chances of corrupting your file system and to make your next powerup be faster by avoiding file system check next powerup. If you have the time to properly shutdown the linux host you will not run the risk, however small, of corrupting your Micro SD card.

Once you have an open SSH console to the Magni, which is explained next on this page, you can shutdown the Magni in a clean way using these commands and actions

If your Magni is a Magni silver and you have the sonar board you can always use the SW1 switch by holding it down for a couple seconds. This will lead to a clean shutdown of the host Raspberry Pi CPU.

If you do not have a sonar board you can wire for yourself a reset switch that is normally open and when connected connects pin 31 of P702 on the MCB to pin 30 of that large 50 pin connector.