-
Notifications
You must be signed in to change notification settings - Fork 41
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
[Discussion] Allow Multi-robot simulation #11
Comments
For now, we see several things that limit/prohibit the usage of the current structure for the multi-robot usecase, e.g.: In robot_state_publisher the hardcoded
from raw3-2 should be replaced by something like
|
The robot_namespace in the gazebo_ros_control plugins for the respective components should be replaced by something like
where
(default value is
|
We need to find a good solution for un-global-namespacing the According to this lines, the Subscribers of the joint_state_publisher can also better be namespaced via the launch file than using a global |
|
I assume you are mainly working within |
The development branches are: The diffs are here: |
As discussed with @ipa-mig @ipa-srd and @ipa-srd-fg we might need to re-think about the topic names from the laser scanner (might need to happen for other sensors too?) Still a problem is within costmap node. After current experiments with raw3-2 are satisfactory, changes need to be applied for all robots, too! |
Since I am not very familiar with git and have been advised to use Currently, I try to apply the changes for all robots. Unfortunately, when there is also a There is also a problem with prosilica. By deleting the leading slashes of the topic names, all topic names are then changed by considering the
|
That's perfect! It's good to have a meaningful and consistent feature branch name in the related repos. As to the robot_description param and the ros_controller: I see your changes a bit sceptical as the the joint_trajectory_controller is widely used in the community and I'd rather not touch it. Maybe it then makes sense to actually have those namespaced robot_description parameters rather than just one global /robot_description. Think about simulating raw3-2 and cob4-2 simultaneously! The robot_descriptions are different then... As to the topic names of the other sensors: We carefully have to look what the names are in the real sensor drivers and then find a consistent solution. We could discuss these issues next week on Monday...? |
+1 to having namespaces |
Sorry for not expressing my ideas in more detail concerning the For discussing the topic issues next week on Monday is suitable for me. The |
@ipa-srd-fg: @ipa-srd @ipa-mig will you join this meeting? |
I'd like to. But cannot say yet due to demo preperations |
Although I won’t have much time today I would join the discussion at least for the main topics. Von: Matthias Gruhler [mailto:notifications@github.com] I'd like to. But cannot say yet due to demo preperations — |
Previous discussion on robot_description namespace: |
As discussed with @ipa-fxm, the work (that we are currently aware of) that still needs to be done in order to allow multi-robot simulation for all robots can be summarized as follows:
|
@ipa-fxm, @ipa-srd-fg Just stumbled upon this migration from tf to tf2 which states that To quote:
I'm not sure whether this actually is a problem or not, but may be worth looking into... |
First PRs have been sent:
|
True! Other repos that definitely need to be considered:
|
The problem with searchParam("robot_description") also occurs when using the rqt_joint_trajectory_controller (see here) And I think we will find it at some more places... |
this is a rqt graph from the current cob4-2 setup. Can be loaded in rqt again (from file). https://gist.github.com/ipa-fmw/323da2de461da730999f |
With the result from the discussion during the ROSCon 2015 Birds-of-a-Feather session the following is agreed:
Therefore, a better approach would be:
First trials with multi-master implementations seem promising and support this approach. Regarding the current set of PullRequests: |
I created a new public repository for temporary experiments regarding multi-robot/mutli-master support: https://github.com/ipa320/cob_multi_robots |
Thanks |
Just for the record: a similar discussion about Multi-Robot namespacing and tf_prefix'ing in gazebo ros-simulation/gazebo_ros_pkgs#333 |
@ipa-srd @ipa-srd-fg
The goal is to prepare the
cob_bringup
/cob_bringup_sim
structure (as well as everything that comes with it, e.g. robot_description, controller,...) to be able to use multiple robots at the same time.This might later be extended to a multi-master setup....
Feel free to add more issues and problems faced in order to be able to solve them in accordance with single-robot usecase
@ipa-fmw @ipa-nhg @ipa-mig FYI
The text was updated successfully, but these errors were encountered: