Questions about using the mocap package

[Vincent1029384756]

Hi, I'm still new to ROS and crazyswarm2. The other day I attempted to use motion capture with Vicon and got really confused.
First, about the marker configurations:

marker_configurations:
      default: # for standard Crazyflie
        offset: [0.0, 0.0, 0.0]
        points:
          p0: [   0.0,    0.0, 0.022]  # top
          p1: [-0.042,  0.042, 0.0  ]  # back left   (M3)
          p2: [-0.042, -0.042, 0.0  ]  # back right  (M2)
          p3: [ 0.042, -0.042, 0.0  ]  # front right (M1)
      default_single_marker:
        offset: [0.0, -0.01, -0.04]
        points:
          p0: [0.0177184,0.0139654,0.0557585]

Do these match the images that are included in the package?

Secondly, about Vicon integration...
I was taught the following workflow when using Vicon in school:

1. Define marker arrangements in the Vicon software.
2. Assign names to each object or robot in the software.
3. Stream tracking data as ROS topics (e.g., /vicon/robot1/pose).
4. Specify in the ROS package which topics to listen to for pose feedback.

However, in Crazyswarm2, it seems the marker arrangement is specified directly in the motion_capture.yaml file instead of the Vicon software. Additionally:

• I don’t see where to specify the topic names for each robot's Vicon data.
• I couldn’t find an explanation in the documentation about how Crazyswarm2 processes Vicon data and determines which data belongs to which robot.

My specific questions

1. How does Crazyswarm2 accept data streamed from Vicon software?
2. How does it differentiate between Vicon data for multiple robots?
3. Is there a way to specify custom topic names for each robot’s Vicon data?
4. Are there any resources or examples that explain how to properly integrate Crazyswarm2 with Vicon?

I’ve already checked the documentation for both Crazyswarm2 and motion_capture_tracking, but couldn’t find answers to these questions.

Any help or pointers in the right direction would be greatly appreciated. Thank you!

[whoenig]

1. It uses the motion_capture_tracking package (which publishes on /poses).
2. There are two modes. You can use the VICON Tracker software, or you use our custom tracker. The custom tracker tracks frame-by-frame, i.e., you need to provide an initial guess of the position of each robot.
3. No, for performance reasons publishing is done via tf nd /poses, only.
4. The documentation for Crazyswarm1. There is also a recent tutorial at https://imrclab.github.io/workshop-aerial-swarms-rss2024/.

[Vincent1029384756]

Thanks a lot! I will try it out today.

[Vincent1029384756]

It worked! Thank you so much for your help.