ROS Resources: Documentation | Support | Discussion Forum | Index | Service Status | ros @ Robotics Stack Exchange
Ask Your Question

use IMU in differential robot?

asked 2019-03-30 09:04:17 -0500

mateusguilherme gravatar image

updated 2019-03-30 11:20:28 -0500


Using an IMU sensor on a differential robot for navigation2d increases accuracy? is it worth?

Since the messages Sensor_msgs/Imu.msg do not consider the magnetic field ...

Has anyone ever made a comparison? any experience report is welcome...


edit retag flag offensive close merge delete


It is a de facto configuration I think..

tianb03 gravatar image tianb03  ( 2019-04-01 23:57:37 -0500 )edit

I do not understand, could you explain better?

mateusguilherme gravatar image mateusguilherme  ( 2019-04-02 08:35:22 -0500 )edit

1 Answer

Sort by ยป oldest newest most voted

answered 2019-04-02 11:50:43 -0500

Yes, an IMU can help you as compared to only wheel encoders! There's a more formal explanation which I think is worth digging into on your own time, but the major point is that an IMU's orientation/gryo drift is only proportional to time where wheel encoders are proportional to distance.

That's a _huge_ result in my opinion making robotics much easier. Time is almost never the enemy but distance is. After moving 1 meter, you may have AMCL, ORB-SLAM, etc update your global position and then your drift is corrected. If you had wheel encoders only then the longer you go without an update the more wacky things are going to get. Versus having an IMU you can trust in the loop, you can now go a very long time without updates assuming practical limits.

edit flag offensive delete link more

Question Tools



Asked: 2019-03-30 09:04:17 -0500

Seen: 346 times

Last updated: Apr 02 '19