DPRG mailing list archives: [DPRG] Re: Feedback on Kip's can-can rules

Sat Sep 22 12:33:05 CDT 2001

I agree with most of the rule changes and much of what Clay
has said.  My comments are dispersed through his email.

First I have a loophole or unexplained item I'd like clarified.

I plan to have my robot move around in area A while trying to 
locate as many of the cans as possible BEFORE leaving area A 
and starting the timer.  Question is: Does the time start when
the robot first moves or when it leaves area A?

> From: Clay Timmons [mailto:ctimmons at cadence.com]
>
> Voice Activation -   I have several reservations about this and
> don't recommend it.   [.snip.]  My recommendation, leave it
> as is with 0-5 and let the judges decide how sophisticated
> the robot starts on command.

I agree with this.  If someone can demonstrate that it truely is 
voice activated would be given the full points.  Perhaps by 
verbally commanding "blink lights", "open gripper", "spin in 
place", (etc.) then "GO".  Crowd noise could be overcome with
a handheld FM microphone and a reciever on the robot.

> Retrieved cans in a pattern - [.snip.]

This is part of the debate from just before the contest.

What is the real goal of the contest?  Is it just to collect the
cans or is it to advance the state of hobby robotics?  Some
have proposed that a giant scoop is a novel or clever or elegant
solution to the problem.  Maybe so.  However, I feel that the
real purpose of our contests is to advance the state of robotics.
A 6 degree of freedom arm that can pick up the cans and place
them in a pattern but misses one of the cans should score higher 
than a scoop that gets all of the cans.

The contest is only a way to score or judge which robot has
advanced the state of hobby robotics the most.  As others have 
said, if your only goal is to collect cans, pay a 5 year old a 
nickle a can...  I believe there should be no way a wall 
follower with a scoop should be able to beat SR04, unless SR04
performs below its usual standard.  (If SR04 only gets 3 or 4
cans and the scoop gets them all, then the scoop should win.)

> Navigation skill - [.snip.] I like leaving the judges
> room to assign points here, 0-10, as it is now for all 4 of our
> RoboRama contests.

I agree (esp. the part about smooth movement).

> Can Collection Skill -  This is the area I think the judges
> should be able to reward the precision 6 axis arm vs. the
> passive scoop and trap method.    With your proposals
> a passive scooper vs. the 6 axis arm both with perfect runs
> would get the same score!  Some would argue that they
> should get the same score because the accomplished the same
> task.   My opinion is that a more robust collector needs to
> be rewarded over less sophisticated method.

See comments under "Retrieved cans" above.

> Knocked down cans left on course -  I argeee with
> your 5 point penatly per can.  I would put this as
> seperate scoring item not can collection skill.   Overall
> I'm hesitant to add scoring items but this is a simple one
> for the judges to deal with.

Yes.  And if the 6 axis arm is able to upright a knocked over
can, then the crowd (and the judges) should "go wild".

> Thanks for typing up your rule proposals.  It is always a challenge
> to get rules updated since you will never get complete consensus
> from club members.

ABSOLUTELY!!!  Thanks to Kip for being willing to step into the 
limelight on this one.

				- Dan Miner