So I looked into the possibilites regarding simulation for a bit
and found vismach [1] and openrave [2].

I couldn't really get used to vismach.
Why are the models e.g. puma560 built as scale models?
It doesn't support collision detection, right?

Also I don't like that the model scripts (python) don't use the parameters
from the LinucCNC config files, but instead need to be hardcoded and don't
reside in the machine config folder.

When homing e.g. the puma560 multiple times the 5th /B axis ends up in
different positions - why?

^^ so vismach seems to be only a quick hack that I suppose is only intended
to give LinuxCNC the possibility for some tests without an actual machine?

Openrave on the other hand looks really promising - but is unfortunately
only
working with the unmaintained linuxcnc_gen_c branch of ftkalcevic.

Points I'm missing about Vismach / Openrave?
Did I miss other solutions?

I don't really like either of those and started working on
getting robotics library [3] running as a simulator for LinuxCNC
through boost.python. It's not yet working but I'm making progress.

Once the simulator is running I plan on using it for gcode verification
(collision detection, speed/acceleration/angular limits) - since I would
like to minimize crashes / other problems with the real robot.

Maybe even path planning is doable and implementing missing PTP features of
LInuxCNC.

What do you think about roboticslibrary + LinuxCNC?

Another possibility could be Gazebo [4] from ros [5] but I have only just
built an urdf model and not really put any more effort into it since.


[1] http://wiki.linuxcnc.org/cgi-bin/wiki.pl?Vismach
[2] http://wiki.linuxcnc.org/cgi-bin/wiki.pl?Ikfast
[3] http://www.roboticslibrary.org/
[4] http://gazebosim.org/
[5] http://www.ros.org/
------------------------------------------------------------------------------
One dashboard for servers and applications across Physical-Virtual-Cloud 
Widest out-of-the-box monitoring support with 50+ applications
Performance metrics, stats and reports that give you Actionable Insights
Deep dive visibility with transaction tracing using APM Insight.
http://ad.doubleclick.net/ddm/clk/290420510;117567292;y
_______________________________________________
Emc-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/emc-users

Reply via email to