Robotics StackExchange | Archived questions

TF_OLD_DATA error when using gmapping with V-REP

Hi,

I'm new to ROS (Kinetic) and dealing with a simulated robot (robotino) and SLAM. The problem is that when i try to launch gmapping by the command:

rosrun gmapping slam_gmapping scan:=/base_scan

I get the following error:

Warning: TF_OLD_DATA ignoring data from the past for frame odom at time 1.53886e+09 according to authority unknown_publisher
Possible reasons are listed at http://wiki.ros.org/tf/Errors%20explained
         at line 277 in /tmp/binarydeb/ros-kinetic-tf2-0.5.16/src/buffer_core.cp

I've already tried to send a empty message to the topic /reset_time but it didn't work.

I'm in a Ubuntu 16.04 and using the SICKTIM310fast laser in V-REP and I publish it's data using the following piece of Lua script: 

laserInfo = {}
intensities = {}
data=sim.getStringSignal("measuredDataAtThisTime")
if data then
    measuredData=sim.unpackFloatTable(data)
    laserInfo['header'] = {seq=0,stamp=simROS.getTime(), frame_id="laser_lixo"}
    laserInfo['angle_min'] = -1.54
    laserInfo['angle_max'] = 1.54
    laserInfo['angle_increment'] = 0.017444444444444
    laserInfo['time_increment'] = 0.0666666667
    laserInfo['scan_time'] = 1.0
    laserInfo['range_max'] = 4.0
    laserInfo['range_min'] = 0.05
    laserInfo['ranges'] = measuredData
    for k=0,table.getn(measuredData) do table.insert(intensities, 0) end
    laserInfo['intensities'] = intensities
    simROS.publish(laserPub, laserInfo)
end

This is what i current get from the topic /tf: https://drive.google.com/file/d/1HTH1x4R0Egs8_Nn9gYW_-rBTKVk35Di2/view?usp=sharing

After i run gmapping a new /tf_static topic is created but i get nothing from it.

This is my current tf tree: https://drive.google.com/file/d/1vrI6bRmO4dcXt_LZuwyMF9rqosnooEr8/view?usp=sharing

Thanks for your help and if i'm missing some information that could be useful please tell me.

Asked by GuiNar on 2018-10-06 17:31:01 UTC

Comments

Is use_sim_time set on the machine? If not, gmapping will try to use normal wall clock time but V-REP might be publishing simulation time (ie: starts from 0). That is not going to work.

You'll have to add more information about your setup to your question for others to be able to help you.

Asked by gvdhoorn on 2018-10-08 01:17:19 UTC

Yes, it's set, if i don't do it i just get a seg fault error. I will try to add some more information but i don't exactly know what is important in this case, could you tell me? Thanks for your help.

Asked by GuiNar on 2018-10-08 08:19:05 UTC

This looks possibly related to: #q304824.

Asked by tfoote on 2018-10-09 18:44:51 UTC

Could you make a minimal example of how to reproduce this error? I've seen several related reports recently but cannot reproduce it. Others have reported that the published tf timestamps are always at the same value.

Asked by tfoote on 2018-10-10 18:36:59 UTC

I can't be sure this is related, but I recently (ie: yesterday) ran into a similar persistent TF_OLD_DATA situation which turned out to be an ABI break. I had a workspace compiled from before the recent ABI breakage (with rostime) and tried to use it on an updated system. All my Gazebo ..

Asked by gvdhoorn on 2018-10-11 01:30:20 UTC

.. plugins broke and I was confronted with a wall of TF_OLD_DATA errors. Nothing worked until I rebuilt my workspace.

Asked by gvdhoorn on 2018-10-11 01:30:44 UTC

To reproduce this i just ran roscore, vrep and then tried to run gmapping, nothing else. I added an image of what i get from the /tf topic and an image of my current tf tree. I also rebuilt my workspace but it keeps getting the same error. Thanks again for the answers.

Asked by GuiNar on 2018-10-13 09:27:50 UTC

Answers

The reason I posted this was to show that the issue was not related to using a simulation. The real answer was that some update and ABI break (i think to rostime) broke my system. I ran apt-get update then catin_make clean. I had to fix a bunch of issues until that completed correctly. Then I ran catkin_make and that error no longer occurred.

I am getting the same error only I am not trying to run a simulation. I have 2 launch files. minimal_baselink.launch which publishes odom and subscrbes to cmd_vel <!-- motorcontrol with correct base frame--> and bigbotnavslam.launch which handles the URDF,RVIZ, move_base, gmapping, and the lidar. <?xml version="1.0" ?> <!-- description and rviz --> <!--- Run gmapping --> <!--- Run Move Base -->

Note use_sim_time is set to false in both.

here are some errors from my log (not in order)

1544295978.191982261 WARN [/opt/ros/kinetic/include/tf/message_filter.h:496(MessageFilter::checkFailures) [topics: /rosout, /tf, /slam_gmapping/entropy, /map, /map_metadata] MessageFilter [target=odom ]: Dropped 100.00% of messages so far. Please turn the [ros.gmapping.message_notifier] rosconsole logger to DEBUG for more information.

1544295978.192033769 WARN [/opt/ros/kinetic/include/tf/message_filter.h:501(MessageFilter::checkFailures) [topics: /rosout, /tf, /slam_gmapping/entropy, /map, /map_metadata] MessageFilter [target=odom ]: The majority of dropped messages were due to messages growing older than the TF cache time. The last message's timestamp was: 1544295977.975927, and the last frame_id was: laser_frame

1544295843.425278847 WARN [/tmp/binarydeb/ros-kinetic-kdl-parser-1.12.11/src/kdl_parser.cpp:208(treeFromUrdfModel) [topics: /rosout] The root link base_link has an inertia specified in the URDF, but KDL does not support a root link with an inertia. As a workaround, you can add an extra dummy link to your URDF.

1544296003.921623294 WARN [/tmp/binarydeb/ros-kinetic-costmap-2d-1.14.4/src/costmap_2d_ros.cpp:109(Costmap2DROS::Costmap2DROS) [topics: /rosout, /cmd_vel, /move_base/current_goal, /move_base/goal] Timed out waiting for transform from base_link to map to become available before running costmap, tf error: Lookup would require extrapolation into the past. Requested time 1544296003.923271802 but the earliest data is at time 3532084496.000032766, when looking up transform from frame [base_link] to frame [map]. canTransform returned after 0.101056 timeout was 0.1.

1544296195.027000729 WARN [/tmp/binarydeb/ros-kinetic-costmap-2d-1.14.4/src/costmap_2d_ros.cpp:109(Costmap2DROS::Costmap2DROS) [topics: /rosout, /cmd_vel, /move_base/current_goal, /move_base/goal] Timed out waiting for transform from base_link to map to become available before running costmap, tf error: Lookup would require extrapolation into the past. Requested time 1544296195.073266109 but the earliest data is at time 3532084496.000032766, when looking up transform from frame [base_link] to frame [map]. canTransform returned after 0.100911 timeout was 0.1.

1544296196.828247318 WARN [/tmp/binarydeb/ros-kinetic-tf2-0.5.19/src/buffer_core.cpp:277() [topics: /rosout, /initialpose, /move_base_simple/goal, /clicked_point] TF_OLD_DATA ignoring data from the past for frame base_link at time 7.08722e+06 according to authority unknown_publisher Possible reasons are listed at http://wiki.ros.org/tf/Errors%20explained

Not sure if my launch includes need to be rearranged because rviz comes up first and it take a while for the lidar to get started. So some of these messages might relate to that. Would it help to see the launch files for the urdf,movebase and gmapping?

Asked by ed on 2018-12-08 21:15:39 UTC

Comments

Please don't use an answer to ask a question. This isn't a forum. Please create a new question and reference this question instead.

Asked by jayess on 2018-12-08 22:24:20 UTC