Angular Twist Covariance always 0

[TheDelus]

Hello,

I noticed that in the odometry msgs the covariance for the angular twist is always 0. Why?

header:
  stamp:
    sec: 1737039030
    nanosec: 30000000
  frame_id: FP_ENU0
child_frame_id: FP_POI
pose:
  pose:
    position:
      x: 647.5685
      y: 183.7218
      z: 0.3091
    orientation:
      x: -0.999542
      y: 0.021723
      z: 0.019087
      w: -0.008953
  covariance:
  - 0.0021
  - -0.0
  - 0.00055
  - 0.0
  - 0.0
  - 0.0
  - -0.0
  - 0.00166
  - 6.0e-05
  - 0.0
  - 0.0
  - 0.0
  - 0.00055
  - 6.0e-05
  - 0.00271
  - 0.0
  - 0.0
  - 0.0
  - 0.0
  - 0.0
  - 0.0
  - 0.00058
  - -8.0e-05
  - 0.0
  - 0.0
  - 0.0
  - 0.0
  - -8.0e-05
  - 0.0001
  - 1.0e-05
  - 0.0
  - 0.0
  - 0.0
  - 0.0
  - 1.0e-05
  - 7.0e-05
twist:
  twist:
    linear:
      x: -6.1946
      y: 0.031
      z: 0.3118
    angular:
      x: -0.02149
      y: 0.0275
      z: -0.01059
  covariance:
  - 0.00697
  - 4.0e-05
  - -0.00104
  - 0.0
  - 0.0
  - 0.0
  - 4.0e-05
  - 0.00535
  - -0.00022
  - 0.0
  - 0.0
  - 0.0
  - -0.00104
  - -0.00022
  - 0.00533
  - 0.0
  - 0.0
  - 0.0
  - 0.0
  - 0.0
  - 0.0
  - 0.0
  - 0.0
  - 0.0
  - 0.0
  - 0.0
  - 0.0
  - 0.0
  - 0.0
  - 0.0
  - 0.0
  - 0.0
  - 0.0
  - 0.0
  - 0.0
  - 0.0

[SN00R]

Dear @TheDelus,

The example shows a message published on the /fixposition/fpa/odometry topic. Therefore, this information comes from the FP_A-ODOMETRY message. In our web documentation for this message and in our covariances documentation we see that only the covariance of the linear velocity is estimated by our system. Therefore, the covariance matrix for the velocity is a 3x3 matrix.

Can you elaborate on your use case of the covariances for the angular velocities?

Kindly reach out to [email protected] for more questions.