IKFast on the Motoman SDA10F

classic Classic list List threaded Threaded
1 message Options
Reply | Threaded
Open this post in threaded view

IKFast on the Motoman SDA10F

Alex Henning-2

I'm trying to run IKFast on the Motoman SDA10F. I've converted the URDF
from the ROS-Industrial repository[1] to a collada file
(attached). Running IKFast on it doesn't seem to work, it freezes after
finding three nonconsecutive axes. I've tried rounding it as suggested
in the MoveIt documentation[2]. That only helps when I round it down to
one decimal place, which produces unacceptable results. When running it
rounded to five decimal places, I get the output seen in
ikfast-61.log. I have to kill it when it hits ~80% of 8GB of RAM after
two hours of it printing nothing and providing no feedback.

I tried updating to the version of IKFast in the master branch on
github. This seems to print more and use less RAM, however it eventually
crashes with "ValueError: A Matrix must have non-zero determinant to
invert." The full output is in ikfast-master.log.

Any help producing IKFast solutions for this robot would be greatly

[1] Motoman Repository: https://github.com/ros-industrial/motoman
[2] MoveIt IKFast documentation: http://docs.ros.org/hydro/api/moveit_ikfast/html/doc/ikfast_tutorial.html

Thanks in advance,
Alex Henning

Dive into the World of Parallel Programming. The Go Parallel Website,
sponsored by Intel and developed in partnership with Slashdot Media, is your
hub for all things parallel software development, from weekly thought
leadership blogs to news, videos, case studies, tutorials and more. Take a
look and join the conversation now. http://goparallel.sourceforge.net/
Openrave-users mailing list
[hidden email]

sda10f.dae (114K) Download Attachment
sda10f.rounded.dae (114K) Download Attachment
ikfast-61.log (3K) Download Attachment
ikfast-master.log (52K) Download Attachment