| ROS Resources: Documentation | Support | Discussion Forum | Index | Service Status | ros @ Robotics Stack Exchange |
I'm using hector GPS in my robot. I followed the instructions to introduce it in the EKF (robot_pose_ekf). However, the estimated pose is much worse taking the gps into account than that obtained ignoring it.
I think that the problem comes from de gps I am using because when I move the robot around it accumulates an error which never disappears. My GPS is Hector gps
Integration of GPS with robot_pose_ekf proved to be difficult for me when I tried it as well. Someone may have a good answer for you, but if it proves too difficult, I recently released the robot_localization package, which contains both an EKF localization node and a UTM transform node. The former can be used in place of robot_pose_ekf, and the latter allows for integration of GPS data. The tutorials page contains a tutorial that details how to integrate a GPS sensor.
[ERROR] [1403082870.757635304, 20.409000000]: Odometry pose transform error: Lookup would require extrapolation into the future. Requested time 20.409000000 but the latest data is at time 20.402000000, when looking up transform from frame [base_footprint] to frame [base_link]
Are all the messages originating on the same machine? At what rate is your base_link->base_footprint transform being broadcast? Feel free to send me an e-mail (see my GitHub page). I may have to have a look at your bag file and then we can update this question when we figure out the issue.
Asked: 2014-06-16 06:10:06 -0500
Seen: 1,186 times
Last updated: Jun 18 '14
ROS Answers is licensed under Creative Commons Attribution 3.0 Content on this site is licensed under a Creative Commons Attribution Share Alike 3.0 license.
