ROS Resources: Documentation | Support | Discussion Forum | Index | Service Status | Q&A answers.ros.org |
I think navfn is quite simple in that it only plans for circular robots. Your's seems a bit lengthy and rectangular. This would lead to a circumference that just doesn't fit through the passage.
Please start posting anonymously - your entry will be published after you log in or create a new account.
Asked: 2012-03-13 06:15:12 -0600
Seen: 186 times
Last updated: Mar 13 '12
How does Dijkstra's algorithms work inside navfn?
Testing performance of my global planner
Navigation path planning over long distances
allow_unknown parameter ignored
Global Planner Parameters in Navigation
Why does the navfn plan a path through a wall
NavFn setting up cost array (costarr) from ROS's costmap
How to get information from the navfn planner execution
what is the reasoning behind current architecture of navigation stack (move_base), the way it is?
questions about dijkstra algorithm used in navfn package to path planning
I am seeing the same behavior, but I can't post a screenshot. In tight areas, the plan starts right after the tight area instead of starting from the robot. Edit: Increasing cost_scaling_factor from 0.1 to 10 fixed this problem for me.