ROS Resources: Documentation | Support | Discussion Forum | Index | Service Status | ros @ Robotics Stack Exchange
Ask Your Question
1

Pioneer 3DX strange sonar behavior

asked 2013-05-31 05:46:54 -0500

JSantos gravatar image

updated 2014-01-28 17:16:42 -0500

ngrennan gravatar image

Hi,

I'm doing experiments with the frontal sonar ring present in the Pioneer 3DX using ROSARIA with ROS Fuerte. However, the output from the array wasn't what I expected. The sonar readings often return the maximum range of the sonar, even though the wall is clearly not so far away. I made a video where is possible to compare the sonar readings with an LFR in my campus corridors. Also, I have uploaded a photo of the corridor to give a general overview of the scenario.

Video

Campus

Yellow dots are the sonar readings (pointcloud published by ROSARIA), red dots are LFR scans and the green dots can be ignored.

I've tried both ROSARIA and the p2os drivers. Both gave me the same output. I also tried it with two more pioneers available in my lab, but they gave me the same results. Is there an hardware issue with the p3dx sonars? Does anyone know if changing the frequency of the sonar readings solves this issue?

Best regards,

João Santos

edit retag flag offensive close merge delete

Comments

I have worked with ARIA libraries, and have no problems with sonars... so I dont think this could be a hardware issue.

gustavo.velascoh gravatar image gustavo.velascoh  ( 2013-05-31 10:49:14 -0500 )edit
1

Check the max range of sonar in the code. Just to confirm whether sonar is working fine try getting some obstacle very close to sonar and see whether it detects it and slowly move it away from robot to get some rough estimate of range of sonar.

ayush_dewan gravatar image ayush_dewan  ( 2013-06-01 07:56:28 -0500 )edit

1 Answer

Sort by » oldest newest most voted
2

answered 2013-06-01 11:38:23 -0500

That looks pretty much like what I would expect from sonar sensors. There's a reason why they are cheap and why LIDARs are generally used when one wants to have (more) accurate data.

It clearly looks like the sensors don't like it when their emitted sound hits walls at oblique angles (as this means it is reflected away from the sensor, thus giving no return). Observe that the two sensors to the sides work reliably in your video and that these are the only ones that consistently emit at a right angle against the walls, thus retrieving strong returns. The material the environment is made of of course also plays a significant role in overall performance.

edit flag offensive delete link more

Question Tools

1 follower

Stats

Asked: 2013-05-31 05:46:54 -0500

Seen: 1,662 times

Last updated: Jun 01 '13