Ask Your Question

Odometry isn't reasonable for rotational motion

asked 2016-02-08 09:21:58 -0600

Orhan gravatar image

updated 2018-03-26 12:19:00 -0600

The thing is; following screenshot is from simulation, on Gazebo. (And Real robot does this too.) ekf_localization_node uses both of imu and odometry. And using only odometry causes this problem too.

I'm following this guide to tune my robot : ROS Wiki Navigation Tuning Guide I've set Decay Time to 30 and started inplace rotation.

Here is Rviz screenshot: Rviz

Odometry isn't reasonable for rotational motion.

Robot's translational motion is good and reasonable. But rotational motion is badly enough for localization. What can I do for correcting odometry data? Or Are there any new issues on ekf_localization?

Thank you.

edit retag flag offensive close merge delete

2 Answers

Sort by » oldest newest most voted

answered 2016-02-09 03:27:48 -0600

Orhan gravatar image

updated 2018-03-26 12:20:32 -0600

Ok, The problem was wheel_width parameter in the urdf (diff_drive_controller is attempting to read this parameter from urdf). It was 0.05 and It caused two contacts under wheel: image description

I've set to 0.001 and problem solved, now odometry looks pretty fine.

edit flag offensive delete link more

answered 2016-02-09 01:30:31 -0600

Procópio gravatar image

for a real robot, I would say what causes this is a wrong wheel size and wheels distance (assuming a diff drive robot), but as the same happens in simulation, I am not sure about the reason.

edit flag offensive delete link more

Your Answer

Please start posting anonymously - your entry will be published after you log in or create a new account.

Add Answer

Question Tools



Asked: 2016-02-08 09:21:58 -0600

Seen: 420 times

Last updated: Mar 26 '18