DPRG List  

[DPRG] Work with researchers?

Subject: [DPRG] Work with researchers?
From: Randy M. Dumse rmd at newmicros.com
Date: Tue Jan 5 20:14:23 CST 2010

Hi gang, toss one out here, some blue sky thinking, here.
Inspired by listening to David Hanson's interview on KERA.

Do you suppose there's any purpose in trying to align hbbiest
with robotics researchers? 

I know we've all got our heros in Robotics. We read about their
work, and think these guys are on an interesting path, and we
might even try to emulate their work. Many of these are in
academia, and so they have a supply of students to work on their
projects. But what they don't have are steady workers who stick
with them more than a semester, to a few years, before they
graduate and move on. Well, that keeps fresh blood rolling in.
But it also may mean people with developed skills might not be
available, and they have to keep "rebuilding their human tools"
every time they get a new, unexperienced student. 

I guarantee you, every one of these researchers will have some
odd ball idea or other for robotics projects they'd like to
explore if only they had the time and personnel to put on it. 

Now it may be they'd not be receptive to outside help for a
number of reasons. Primary being it if took more of their time
than it saved. But there are other reasons. Lack of control.
Lack of funding. Lack of availability of critical equipment.
Lack of time to give guidance. But it also might be, they'd love
a volunteer hand, and have never considered asking. 

To give myself as an example, my students did a bit of playing
with the CMUCam. They got a sumo that tracked a color. But the
search algorythm only turned one direction if it lost the
object. I'd love to see that algorythm extended. I have one of
those expensive IMU's dpa used in his robots. I'd love to see
the software interface worked out. I posted about my "Mind over
Matter" experiment ideas, where we take two identical sumos,
give one a physical advantage and test to prove it became
superior, then give the other a programatic change and see if it
could overcome a physical disadvantage. That would prove what we
all suspect, but no one has ever proved, that "mind" alone in a
robot can overcome "matter". It's one of those ideas that might
lead to a pivotal paper oft quoted in the industry.

So if we had a mechanism of hooking up dedicated hobbiest who
like to tinker anyway, and would like some ideas to work on,
with reseachers who might be able to use the help, and maybe
offer some credit in the literature for their volunteers... Does
that sound like a useful path anyone would like to pursue?


More information about the DPRG mailing list