LiDAR used for the elevation map

[ylenianistico]

Hello and congrats for this work!

I have a question regarding the LiDAR used for the elevation map and to get the odometry.
Did you use the built-in Go2 LiDAR, or did you mount an additional one?

Thanks in advance

[NtagkasAlex]

Hello @ylenianistico and thanks for your interest,

Exactly, I used the built-in Go2 LiDAR (4D L1 LiDAR). However, due to its relatively low sampling rate, using a Livox-360 would be more suitable for generating a denser elevation map.

Adapting to a new LiDAR should be fairly straightforward, as it would mainly require changes to the transform_sensors package.

[ylenianistico]

@NtagkasAlex Thanks!

I'm asking because we're also working with a Go2 robot, but we're experiencing several issues with the 4D L1 LiDAR, in particular, the IMU of the LiDAR (so NOT the main IMU of the robot) is outputting bad values.
To read the topics we're following this.

I have the feeling that the problem is related to some updates to download, but I'm not able to figure it out. Have you ever experienced something similar?

[NtagkasAlex]

Hello @ylenianistico ,
I didn’t encounter any specific issues with the quality of the LiDAR IMU measurements on our setup. I was subscribing to the /utlidar/imu topic, which by default publishes the IMU data from the 4D L1 LiDAR, and the values looked consistent.

Could you share a bit more detail about the problem you’re seeing (e.g., the type of anomalies in the IMU data, whether they are noise, jumps, or timing issues)? That might help narrow down whether the root cause is more likely hardware- or software-related. It’s also possible that Point-LIO requires some additional tuning to work reliably with your configuration.

Let me know, and I’ll try to help further.

[ylenianistico]

Hi @NtagkasAlex thanks a lot for your reply.

It seems that there might be some problems with our Go2 LiDAR, then.
Basically, from the topic /utlidar/imu, imu data look corrupted, in particular the linear acceleration. Here an example of what we're getting:

ros2 topic echo /utlidar/imu --once gives:

header:
  stamp:
    sec: 1765278830
    nanosec: 770655393
  frame_id: utlidar_imu
orientation:
  x: -5.837657760449087e-33
  y: -5.8376460055055784e-33
  z: -5.837656291081148e-33
  w: -0.9983124136924744
orientation_covariance:
- 0.0
- 0.0
- 0.0
- 0.0
- 0.0
- 0.0
- 0.0
- 0.0
- 0.0
angular_velocity:
  x: -0.001274624839425087
  y: 0.13813883066177368
  z: 0.00012228553532622755
angular_velocity_covariance:
- 0.0
- 0.0
- 0.0
- 0.0
- 0.0
- 0.0
- 0.0
- 0.0
- 0.0
linear_acceleration:
  x: -2.1734925204012225e+24
  y: -2.9069865676201995e+28
  z: 2.0209584469453046e+18
linear_acceleration_covariance:
- 0.0
- 0.0
- 0.0
- 0.0
- 0.0
- 0.0
- 0.0
- 0.0
- 0.0

I'm not understanding if the sensor itself could be broken, or if the software needs to be update.

We were also thinking of trying to use the main imu of the robot, instead of the L1 one, but there are some other problems with the timestamp, so synchronization with the LiDAR's point cloud could be problematic.