Great introduction to tf, thanks for that. The author makes an error saying at 10:51 that the "publishing occurs at 100 Hz"; actually the 100 represents milliseconds, meaning 10 Hz, and this is what we see next on the tf tree: "Average rate: 10.355". Note the command written on the screen indicates correctly "period_in_ms".
Many thanks for your efforts. Please, when I teleop my turtlebot3 (without any SLAMing, just navigate it manualy) and record data in a bag file, I see that /odometry topic carried the robot position relative to /odom frame Which is the same data of transformation between /base_link and /odom frames. My questions is : 1. Is this a filtered (/estimated) position ? 2. Is it a ground truth data ? 3. What is the relation between this situation and deadreckoning algorithm ? 4. (When I run GmappingSLAM) is the transformation data between (/base_link) and (/map) frame, the same as data of topic odmetry mentioned above !!! In summary , which one is the ground truth and which one is the estimated position ?
Great questions . Before answering these questions I would like you have your input on is there a GPS in turtlebot3 in webots which is giving you data or is it wheel position sensor which is giving you data ?
@@coolrobotics Thanks for your reply. It is turtlebot3 that give its pose as /odom topic data or /footprint frame relative to /odom frame in /tf topic. Which is (as I knew ) encoders and IMU data (without filtering of fusion technique ) .
If you want to use the sendBroadcaster use it as : joint between -> grand-parent parent and second joint between grand parent child. it is possible but you will have to individually control them.
In ouder to prove use of TF Think about ROS without TF. how will you do forward kinamatics? For example , How will you estimate where is an end effector if in 3 axis robot. axis 1 : 30 degrees axis 2 : 45 degrees axis 1 : 60 degrees you will have to manually do lot of math.
Good short overview, thank you!
Glad you liked it!
Great introduction to tf, thanks for that. The author makes an error saying at 10:51 that the "publishing occurs at 100 Hz"; actually the 100 represents milliseconds, meaning 10 Hz, and this is what we see next on the tf tree: "Average rate: 10.355". Note the command written on the screen indicates correctly "period_in_ms".
You are right…
Thanks for the note…
👍🏽
Thank you for your tutorial. This was the explanation I needed to overcome my tf issues. Will know soon if I’m successful!! Subscribed!
nice intro, thank you for this :)
Glad you like it
Thank you
You're welcome
Many thanks for your efforts.
Please, when I teleop my turtlebot3 (without any SLAMing, just navigate it manualy) and record data in a bag file, I see that /odometry topic carried the robot position relative to /odom frame
Which is the same data of transformation between /base_link and /odom frames.
My questions is :
1. Is this a filtered (/estimated) position ?
2. Is it a ground truth data ?
3. What is the relation between this situation and deadreckoning algorithm ?
4. (When I run GmappingSLAM) is the transformation data between (/base_link) and (/map) frame, the same as data of topic odmetry mentioned above !!!
In summary , which one is the ground truth and which one is the estimated position ?
Great questions .
Before answering these questions I would like you have your input on is there a GPS in turtlebot3 in webots which is giving you data or is it wheel position sensor which is giving you data ?
@@coolrobotics Thanks for your reply.
It is turtlebot3 that give its pose as /odom topic data or /footprint frame relative to /odom frame in /tf topic.
Which is (as I knew ) encoders and IMU data (without filtering of fusion technique ) .
Turtlebot3 burger (practical implementation)
Thank you for the tutorial ^_^ Could we use the sendBroadcaster between child and grand-parent to control a set of joints?
If you want to use the sendBroadcaster use it as : joint between -> grand-parent parent
and second joint between grand parent child.
it is possible but you will have to individually control them.
Thank you very much !
If we can't control a set of joints, what is the utility of TF?
In ouder to prove use of TF
Think about ROS without TF.
how will you do forward kinamatics?
For example , How will you estimate where is an end effector if in 3 axis robot.
axis 1 : 30 degrees
axis 2 : 45 degrees
axis 1 : 60 degrees
you will have to manually do lot of math.
yes, you are right, but it would be nice if we can control a set of joints by using the TF tree ^_^
Yess 👍🏽
sir can we move tf1 and tf2 parallelly by scripting python code in ROS2?
Yes you can.
Because the TF2 implementation also considers legacy as well.
@@coolrobotics sir please make separate video for this.
Sure let’s look into our video schedule.
Thanks for the suggestion 😇
@@coolrobotics Sir we haven't get leader follower formation in ros2 video yet
Check other comments 😇
how do i do this for two drones
ruclips.net/video/UWsya46ZG4M/видео.html&ab_channel=IntelligentQuads
this is what i currently have implemented
Namespace your 2 drone URDF and you should be good to go
Let us have a look at the video.
Will get back to you soon 👍🏽
@@coolrobotics thank you i would really appreciate since im new to ros and would like implement TF to the example shown in that video
gazebo iris will already have its URDF. If not you can make a urdf of drone and load it in your ROS system.
First view😅
Thank you for your support