Yvonnn's profile - activity

Hello,

In this video from Tom Moore at time 12:36, I can see that he sets IMU yaw to true. The consequence of this is that as soon as the robot is started, there is an non zero angle between base_link and map, with X axis of map pointing to the east (except if robot is facing east of course). I think there is an inconsistency with what this picture says . I understand here that map and base_link should be aligned at start so that navsat_transform_node can properly compute a transform between utm and map, thanks to the theta angle given by the IMU. Having map and base_link aligned is mandatory because the IMU reports the angle between base_link and east.

Can you please point me my mistake, and if there is no mistake, how should map and base_link be related when using navsat_transform_node ?

Thank you

Yvon

Hello,

I am fusing odometry from a lidar (I have it thanks to undocumented pub_odometry parameter of hector_slam), IMU and GPS data. I am using the dual_ekf_navsat_example. My odom message has no twist information, only pose. So that it is taken into account by ekf_se_odom, I need to publish it with frame_id = odom, because world_map of ekf_se_odom is odom. Then, if I want my odom message to be fused again in ekf_se_map, I need to publish it with frame_id = map, because world_map of ekf_se_map is map. So... I guess I have to duplicate my odom message... am I right or am I missing something ?

Yvon

Well, I've just checked it one more time : pose with frame_id = odom not used by ekf_se_map with world_frame = map but u

When using navsat_transform_node, isn't it a problem to use IMU yaw to compute transform between map and base_link ? Hel

How can navsat_transform_node deal with base_link and map not aligned when GPS data is received ? Hello, I do not unders

How can navsat_transform_node deal with base_link and map not aligned when GPS data is received ? Hello, I do not unders

This : "world-fixed reference frames like odom and map are always orientated so that the 'Y' axis faces magnetic north"

This : "world-fixed reference frames like odom and map are always orientated so that the 'Y' axis faces magnetic north"

Not sure you got it. My pb is strongly linked to this simpler question : will an instance of robot_state_estimator with

What is good practice to fuse pose from a unique odom message in robot_localization in both ekf_se_odom and ekf_se_map ?