mutual robot recognition for collision avoidance in navigation
Hi, I am posing this question since I haven't found a complete answer anywhere else. I'll briefly go over what are my needs.
I need two Turtlebot3 to recognize each other as an obstacle during navigation so that one doesn't crash onto another. With the standard configuration of the navigation stack, that doesn't come out of the box.
So it led me into thinking two possible solutions:
- add a costmap layer to my navigation stack. Unfortunately I understand that it's only extensible via C++ language and since I don't know much of it and, for the purpose of my project, don't intend to learn it, I see this as a possible but less viable solution
- using a computer vision package that allows me to recognize a robot by its shape. This, however, lets me recognize it but not avoid it during the computation of a navigation plan. If I'm mistaken and there's a way to do it, please let me know.
I don't know if there are any other possible solutions to this problem, these are the only that come to my mind.
EDIT: I had to clarify: I am using the standard global planner and a DWA planner, my robots are equipped with lasers and I'm using AMCL to estimate their positions. Regarding the map, it's not fixed: may be known, may be unknown aswell (and will therefore launch gmapping
instead of AMCL).
Can you please clarify: Do you have a laser scanner that can detect the other robots? Have you considered sending robots localization poses to each other?
Yes I have, I'm using a Turtlebot3 so I thought it was implied
And why does it not pick up the other robots in the laser scanner data?
It does, but it doesn't put them (their position) into the costmaps.