Designators in CRAM framework

asked 2012-09-18 22:52:07 -0600

dgerod
113 ●6 ●9 ●15

updated 2012-09-18 22:53:20 -0600

I have just finished reading CRAM core and KnowRob, and I have an overview of the system at this moment.

I have read the following information:

KnowRob wiki (http://www.ros.org/wiki/knowrob)
KnowRob PhD dissertation
Several articles about CRAM, some of them from TUM webpage (http://ias.cs.tum.edu/research/cram)
CRAM information in ROS.org (http://www.ros.org/wiki/cram_core)

However, I still do not understand "designators" in CRAM. How they work? Do you know where I can find more information about them? "Designators" are the module most interesting for me. Thanks.

edit retag flag offensive close merge delete

Comments

@makokal - Thanks for your answer but the link is not working.

dgerod ( 2012-09-19 02:47:11 -0600 )edit

add a comment

3

answered 2012-09-19 05:07:04 -0600

Lorenz

22731 ●129 ●202 ●316 http://ias.in.tum.de/p...

Thanks for you interest in that package. You just made me create some documentation on the wiki page. It should give you a high-level overview of what designators are and how their interface looks like. For examples on how to use them, have a look at the source code of the package cram_plan_library in the stack cram_highlevel. Packages that implement specialized location resolution mechanisms in that stack are:

location_costmap: provides basic functionality to construct costmaps for sampling location designator solutions.
semantic_map_costmap: uses Knowrob's semantic map to generate costmaps for generating poses on, in etc. objects in the semantic map, e.g. on the counter top.
occupancy_grid_costmap: provides a simple implementation for poses to see and to reach specific poses by using the 2d map provided by navigation for inferring locations the robot can navigate to.