0
$\begingroup$

Rosanswers logo

Hello everyone,

I'm currently using move_base and it works fine for me. I'm going to write my own planner based on a real-time planning algorithm. As you know in the real-time search there is no complete path to the goal because we need to decide what action to take soon. So we search forward until s time limit is reached, then the path with minimum cost is selected and this cycle repeats until we reach the goal (there is no guarantee that the path is optimum). So I think I need to wipe out local_planner and global_planner and replace them with only one real-time algorithm. I took a look at the move_base code (https://github.com/ros-planning/navigation) and it looks so confusing to me. I don't know where should I start and what parts of the code should be replaced or removed. I appreciate your help in advance.


Originally posted by Kasra on ROS Answers with karma: 86 on 2014-07-31

Post score: 0


Original comments

Comment by David Lu on 2014-08-01:
What parts of move_base would you like to retain?

Comment by Kasra on 2014-08-01:
I think I can keep cost maps for a real-time planner and only wipe out global and local planners.

$\endgroup$

1 Answer 1

0
$\begingroup$

Rosanswers logo

If all you want to keep is the costmap, I would suggest starting with something simpler like costmap_node or perhaps the global_planner_node.


Originally posted by David Lu with karma: 10932 on 2014-08-01

This answer was ACCEPTED on the original site

Post score: 2


Original comments

Comment by Kasra on 2014-08-01:
I just realized that I need to wipe out also cost maps and write my own environment tracker system. So most of the move_base is gone. I only need to use amcl for localization and then write my own planner. Still I don't have a clear idea of how should I do that! Thanks

Comment by David Lu on 2014-08-01:
AMCL is usually run in a separate node from move_base, so that part should be easy to separate.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.