ROS Team,
I'm having a Nav2 issue.
After I launch my robot, slam
localization, then Nav2, navigation launches ok, then complains:
[lifecycle_manager-10] [INFO] [1760489547.997986714]
[lifecycle_manager_navigation]: Creating bond timer...
[controller_server-1] [INFO] [1760489613.643738544]
[local_costmap.local_costmap]: Message Filter dropping message:
frame 'laser_frame' at time 1760489612.625 for reason 'the
timestamp on the message is earlier than all the data in the
transform cache'
[controller_server-1] [INFO] [1760489613.735305673]
[local_costmap.local_costmap]: Message Filter dropping message:
frame 'laser_frame' at time 1760489612.731 for reason 'the
timestamp on the message is earlier than all the data in the
transform cache'
In Rviz2 the map looks good, the robot's in the right place, but if I try 2D Goal Pose or Nav2 Goal,
The robot avatar doesn't move and the physical robot doesn't try to move.
Navigation fails to plot a path even though the way is clear in the Rviz map.
It will respond to joystick.
I wasted a 4 day weekend on a drunkard's walk with Claude-Sonnet-4.5 trying to troubleshoot this.
His summary is this:
LaserScan timestamps are ~2 seconds in the future, causing transform extrapolation errors in Nav2.
scan.stamp
field is
initialized to 0 (not populated by the driver)node->now()
instead of scan.stamp
still produces timestamps 2 seconds in the futureuse_sim_time
is correctly set to false
on the ydlidar nodedate
)scan.stamp
directly → gave epoch time
(1970)node->now()
→ timestamps still 2
seconds ahead/clock
topic publisher → none existsuse_sim_time
parameter → already falseThe root cause of why node->now()
returns future
time remains unidentified.
Perhaps the ROS Sig might give me their thoughts this evening.
Thanks!
RoverDoc
James H Phelan "Nihil est sine ratione cur potius sit quam non sit" Leibniz
To view this discussion visit https://groups.google.com/d/msgid/hbrobotics/49889669.3044821.1760487833656%40mail.yahoo.com.
Team,
Thanks for the guidance last Tues' ROS meeting to not try to separate my launch file components like "legs of a frog" and expect it to jump. But the problem persists when all are launched from stingray.launch.py.
But although it launches without errors, once it settles I get the persistent error:
[controller_server-7] [INFO] [1760915844.079459033] [local_costmap.local_costmap]: Message Filter dropping message: frame 'laser_frame' at time 1760915842.992 for reason 'the timestamp on the message is earlier than all the data in the transform cache'
The robot responds appropriately to the joystick, but Nav2 fails to even plot a path to Nav2 Goal, let alone attempt to move, instead giving the above timestamp error.
I wasted a 4 day weekend last week with Claude-Sonnet-4.5 on a "drunkard's walk" down many "rabbit holes" (to thoroughly homogenize metaphors) getting bad troubleshooting advice none of which came close to solving the problem. If you want to suffer through that whole dialog, it's here:
https://poe.com/s/ZXUayH9LRfIBGBjHdaNk
Google led me to someone with a similar problem here:
https://answers.ros.org/question/393581/
but I was unable to recognize any actionable advice. Perhaps someone more knowledgeable than I...?
Someone suggested a network timing issue, but all is running on the Pi5 on ROS2 jazzy on Ubuntu 24.04 except Rviz2 and joystick which are running on a Dell CORE i7 laptop also running Ubuntu 24.04 and ROS2 jazzy.
My github is:
https://github.com/JHPHELAN?tab=repositories
articubot_one for the navigation files and
articubot_one/robots/stingray/ for robot specific files
Any insight appreciated!
James
James H Phelan
"Nihil est sine ratione cur potius sit quam non sit" Leibniz
To view this discussion visit https://groups.google.com/d/msgid/hbrobotics/5eeac74e-4cd2-435e-8922-93eeb6ae8f56%40hal-pc.org.
[controller_server-1] [INFO] [1760489613.735305673] [local_costmap.local_costmap]: Message Filter dropping message: frame 'laser_frame' at time 1760489612.731
Pito,
Which "friend" did you ask?
James H Phelan "Nihil est sine ratione cur potius sit quam non sit" Leibniz
On Oct 21, 2025, at 9:27 PM, 'James H Phelan' via HomeBrew Robotics Club <hbrob...@googlegroups.com> wrote:
To view this discussion visit https://groups.google.com/d/msgid/hbrobotics/c74570ae-1f58-4614-b9ee-a10971e742dc%40hal-pc.org.