search

MISTLab / Swarm-SLAM

Sparse Decentralized Collaborative Simultaneous Localization and Mapping Framework for Multi-Robot Systems
https://lajoiepy.github.io/cslam_documentation/html/index.html
MIT License
391 stars 40 forks source link

MarkerArray Status: Error in cslam_visualization on rviz #39

Closed RyanAspen closed 4 months ago

RyanAspen commented 4 months ago

I'm currently running a baseline experiment with just the first ground Graco dataset with Lidar and I'm seeing what may or may not be an error. Running "ros2 launch cslam_experiments graco_lidar.launch.py" in one terminal and "ros2 launch cslam_visualization visualization_lidar.launch.py" in another terminal, I get an updating visualization as expected. However, I get the following on rviz: image. The error messages appear as follows: "poses/1 No transform to fixed frame [robot0_map]. TF error: [Could not find a connection between 'robot0_map' and 'robot1_map' because they are not part of the same tree.Tf has two or more unconnected trees.]"

Is there something more I should be doing to remove this transform error, is this a true error in the source code, or is this not really a problem?

To clarify, I am running on Ubuntu 22.04 with ROS Humble on WSL.

EDIT: Tried this again with 3 datasets instead of one. Still running into the same issues, and the visualization window gets many of these warnings:

[rviz2-2] Warning: TF_OLD_DATA ignoring data from the past for frame velodyne at time 1661304525.900000 according to authority Authority undetectable

lajoiepy commented 4 months ago

Hi! Thanks for your interest in Swarm-SLAM! The TF transform between robot0_map and robot1_map only appears when a inter-robot loop closure is obtained, and pose graph optimization is successful. This may take a while since there needs to be an overlaps between the robot maps, and the system needs to find it through the inter-robot loop closure detection mechanism. Makes sure that your odometry is running ok and makes sense, then try running the full sequences.