|
[DPRG] encouragement
Subject: [DPRG] encouragement
From: dpa
dpa at io.isem.smu.edu
Date: Wed Apr 16 13:51:09 CDT 2008
Howdy,
Eric (Sumner) reports from his visit to Houston last weekend that he
saw three robots capable of successfully completing the DPRG Outdoor
Challenge #1 at next month's gathering. Great news!
He also reported on some concern that robots might not be entered
if their builders felt they were not yet capable enough to compete
effectively, in that and the other challenges.
I'd like to actively encourage anyone who's robot can even remotely
complete the first challenge to do so. Don't be discouraged if your
'bot can't get back any closer than 50 feet to the origin waypoint
(or whatever). Hitting the target accurately is a refinement, but
not necessary. The main goal is to get a robot built and working
to the point where refinements are possible.
And who knows? 50 feet might be good enough to win --- no reason to
assume anyone else is better at this than you are. The robot with
the most expensive sensors is not automatically the best navigator:
they are two different problems.
A good analogy for this is throwing darts at a dart board. I'm no
good at that. But my sensors (eyes in this case) work just fine:
20-20 vision. So the sensors are excellent, as good as the champions.
But the sensors are not the problem. It is the skill in using
those sensors to do fine-scale muscle control that is the problem.
I don't have the skill to control my hand/arm in such a way as to
produce the proper trajectory of the dart through the air in order
to accurately hit the target. It's not a sensing problem, it's a
control problem. The same is true with mobile robots.
Your robot does not have to hit the origin waypoint with any
particular accuracy in order to successfully complete the
challenge. So don't be scared off by the other more difficult
challenges. If each robot that enters can complete challenge #1,
irrespective of the size of the error, then the event will be a
complete success. These are supposed to be incremental challenges,
and the other ones can wait for next time.
But if you can successfully run challenge #1, even if returning no
where near the origin, then the bulk of robot building will be
behind you. And the next three challenges, which really have to
do with adding sensors and writing software, are much easier to
complete.
best regards,
dpa
More information about the DPRG mailing list
|
