Why does TEB plan cross over obstacle inflations about teb_local_planner HOT 2 OPEN

Hi Kevin,

thank you for reporting this issue.
The common inflation layer of the costmap_2d is not taken into account by the planner,
since the internal optimization cannot be applied to grid cells. As an alternative, I added a the possibility to specify a small penalty for robot positions close to obstacles (parameter inflation_dist), but I haven't played a lot with it yet.

Nevertheless, your resulting trajectory in pink das not look well. Do you have further information on how I can reproduce this? By the way, your costmap does not seem to correspond to your static map or amcl failed to localize?

from teb_local_planner.

Hi there! Has there been any progress made on this front?
I have a pretty wide robot that is convinced it can both: fit through tight spaces, and muscle objects out of the way. The desired functionality would be to fail and halt the robot.

Here is a link of a screen cast of the planner operation when an object is introduced. we are using the inflation_dist parameter. I'm happy to provide any other details which could be helpful in diagnosing.
gDrive Link

Also a very big thanks for this awesome planner!

from teb_local_planner.