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

Replace ompl_interface with moveit_ompl_planning_interface #117

Open
mamoll opened this Issue Aug 23, 2016 · 4 comments

Comments

Projects
None yet
3 participants
@mamoll
Contributor

mamoll commented Aug 23, 2016

@ryanluna has written an amazing replacement for ompl_interface. The repo is available here: https://github.com/KavrakiLab/moveit_ompl_planning_interface. He has moved on to Google X and might not have a whole lot of time to help in getting this done. The replacement code allows users to easily add new plugins by just editing a YAML file. This is why you should care:

  • It allows for the easy development of new planners without being forced to recompile MoveIt! or OMPL from scratch.
  • It is also useful for creating customized planning contexts / parameter settings for a particular robot / application.
  • It is stable and usable right now. Several members of our group have been using it for years.

See README.txt in the repo for details. I am not as familiar with the low-level details, but am willing to help as much as I can.

@mamoll

This comment has been minimized.

Show comment
Hide comment
@mamoll

mamoll Aug 24, 2016

Contributor

Perhaps this is not completely obvious, but the moveit_ompl_planning_interface (or OMPL) does not require that a planning algorithm uses sampling, so in principle if someone wanted to create a plugin that implements, say, Multi-Heuristic A*, then this would be much easier with this new plugin system.

Contributor

mamoll commented Aug 24, 2016

Perhaps this is not completely obvious, but the moveit_ompl_planning_interface (or OMPL) does not require that a planning algorithm uses sampling, so in principle if someone wanted to create a plugin that implements, say, Multi-Heuristic A*, then this would be much easier with this new plugin system.

@v4hn

This comment has been minimized.

Show comment
Hide comment
@v4hn

v4hn Sep 26, 2016

Member

@mamoll should we, in your opinion, move the code in with MoveIt, move the repository to the ros-planning organization or officially rely on KavrakiLab?

The module looks intriguing, I fully agree.

Member

v4hn commented Sep 26, 2016

@mamoll should we, in your opinion, move the code in with MoveIt, move the repository to the ros-planning organization or officially rely on KavrakiLab?

The module looks intriguing, I fully agree.

@mamoll

This comment has been minimized.

Show comment
Hide comment
@mamoll

mamoll Sep 26, 2016

Contributor

I am all for integrating this with MoveIt! It'd be nice if Moveit! could be this magical black box that solves all motion planning problems optimally, but in reality people will have to tinker with control knobs or customize planners. What Ryan has written will make this so much easier. I am not core MoveIt! developer, but with some help from Ryan I had no trouble defining new planning contexts for Robonaut2 and customize some planners.

(For future reference, see also discussion on issue #197.)

Contributor

mamoll commented Sep 26, 2016

I am all for integrating this with MoveIt! It'd be nice if Moveit! could be this magical black box that solves all motion planning problems optimally, but in reality people will have to tinker with control knobs or customize planners. What Ryan has written will make this so much easier. I am not core MoveIt! developer, but with some help from Ryan I had no trouble defining new planning contexts for Robonaut2 and customize some planners.

(For future reference, see also discussion on issue #197.)

@davetcoleman

This comment has been minimized.

Show comment
Hide comment
@davetcoleman

davetcoleman Aug 15, 2017

Member

+1 to moving towards this new interface

Member

davetcoleman commented Aug 15, 2017

+1 to moving towards this new interface

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment