Do a circuit loop with a mobile M3 node

Difficulty: Intermediate

Duration: 20 minutes

Prerequisites: Configure SSH Access / Experiment CLI client / Nodes Serial Link Aggregation

Description: The aim of this tutorial is to submit an experiment with a mobile M3 node and interact with it. A mobile node is a node embedded on a robot with predefined trajectories. You will book one mobile M3 node, set it up with a sample firmware file. Once the experiment running, you will interact with the firmware running on the node and graph IMU sensors values. After the experiment, you can plot the mobile node trajectory.
Since robots features are not yet included in the webportal, you will use here the cli-tools.

Launch the experiment

  1. Choose your node:
    • Click on Activity/View Gannt Chart
    • Filter resources by choosing all turtlebot2 in the list
    • Choose a node available and keep its node id
    • Note: Try to take a node which was inactive the last two hours (to be sure to have enough battery).


  2. Choose your circuit, within the list returned by robot-cli:
    robot-cli get -l
  3. Download the simple_robot.elf firmware binary for m3 nodes. For information, you can have a look at the firmware code.
  4. Submit an experiment with cli-tools, either from an SHH frontend or from your computer. In this example, we chose the m3-258 from Lille doing the square1 circuit:
    experiment-cli submit -d 20 -n simple_robot_tutorial -l lille,m3,258,simple_robot.elf,mobility=square1

    Or you may also let the scheduler choose the node just by describing node properties:

    experiment-cli submit -d 20 -n simple_robot_tutorial -l 1,archi=m3:at86rf231+mobile=1+site=lille,simple_robot.elf,mobility=square1

    You get an experiment id.

Interact during experiment

  1. In a terminal on your computer, give access as a tcp socket (explanation in Advanced usage section of Nodes Serial Link Aggregation tutorial) :
    socat tcp-listen:20000,fork,reuseaddr "exec:ssh <login>@<site> 'serial_aggregator'"
  2. In another terminal on your computer, download the script from the our Openlab repository on GitHub.
    Note: you need to have a working Python with Matplotlib installed.
  3. Launch the script with the following parameters. Use the right TCP port configured with socat before. Use the right nodeid (e.g. m3-258) in order to filter incoming traffic from the serial_aggregator. Avalaible parameter for sensor are Gyr (gyrometer), Acc (accelerometer), Mag (magnetometer).
    python ./ --port 20000 --sensor Gyr --nodeid m3-258


Mobile node’s trajectory

  1. Connect to the SSH fronted of Lille with X11 forwarding
    ssh -X <login>
  2. The mobile node trajectory is stored in your home folder in .oml files. You must wait for the end of the experiment to retrieve the file.
    <login>@lille:~$ less ~/.iot-lab/<experiment id>/robot/m3-258.oml
    protocol: 4
    domain: 39738
    start-time: 1459778452
    sender-id: m3-258
    app-name: robot_position_measures
    schema: 0 _experiment_metadata subject:string key:string value:string
    schema: 10 robot_pose timestamp_s:uint32 timestamp_us:uint32 x:double y:double theta:double
    content: text
    10.298170       10      1       1459778462      878753  2.526062        0.129655        -1.563731
    10.398204       10      2       1459778462      978787  2.526062        0.129655        -1.563731
    10.498178       10      3       1459778463      078761  2.526062        0.129655        -1.563731

    Trajectory Measure Table

    Measure Unit
    x, y
    meters, cartesian coordinates
    meters, altitude which is a constant
    radians, orientation around z axis
  3. Plot your file using the oml_plot_tools:
    <login>@lille:~$ plot_oml_traj -i ~/.iot-lab/<experiment id>/robot/m3-258.oml

    You can use --circuit-name and --circuit-file options to observe the mobile node trajectory respectively with the planned circuit and the site map.


Full plot of the mobile node trajectory

Plot of the mobile node trajectory with planned circuit representation

Plot of the mobile node trajectory included into site map

Additional details

During an experiment, a mobile node follows a trajectory chosen by the user (see step 2). Some details about the robots and
the trajectories for each site can be found on the GitHub wiki.

The robot-cli tool is the dedicated client to get the status of a robot or get and update mobilities. See the dedicated tutorial for more information.

It is also possible to visualize the robot trajectory in 3D using Gazebo on your PC workstation. For details, see the GitHub wiki.