I'm trying to use slam_toolbox (2d LIDAR) in conjunction with robot_localization (IMU + Wheel Odom). So far my best idea for managing the frames is as shown in the pictures below. However, this results in the map
frame in Rviz flickering a lot, and printing the frames shows that odomNew
is bypassed. My goal transforms are odom (wheel encoders) <- odomNew (fused with IMU) <- map (slam'd with LIDAR readings). However, I may be approaching this wrong alltogether.
This is my config file for slam_toolbox
This is my config file for robot_localization
Here is my view_frame when I run it:
It basically ignores the odomNew frame alltogether. Is there a better way to do this? Both robot_localization and slam_toolbox work individually, but I can't run both at once properly. Any help is appreciated.
rosrun tf view_frames
, see here? Just to make sure your frames are actually connected as intended. $\endgroup$