DPRG
DPRG List  



[DPRG] Re: Robo-magellan

Subject: [DPRG] Re: Robo-magellan
From: Pete Miles petem at ormondllc.com
Date: Tue Sep 28 19:09:00 CDT 2004

This was a fun contest.  Far more robots competed than we ever anticipated.

I didn't get a chance to see it.  I was too busy trying to find a "lost" hand cart for the Seattle Center.

This event definately showed that there is significant interest in having more challenging contests than the traditional
ones we have.

I would have to say, we really never thought about any form of criteria for determining first, second or third, other
than who completed the course  We offered $1000 for the fasted robot to complete the course.  No robot completed the
course, so no one received the prize.  Out of habit, we made 1st, 2nd, and 3 place plaques for the contest.  Then when
the contest started we soon realized we really didn't have a way to differentiate a 1st, 2nd, or 3rd place when no robot
completes the course.  We were more focused on how to determine the winner when more than one robot completing the
course.  This was a big oversight on our part.  (we have a similar problem with Line Maze, how do you figure out 1st,
2nd, 3rd when no one completes the course).

The easiest thing to do was to say, "well no one completed the course, so see ya next year", and pitch the plaques into
the garbage, or we come up with someway to award the robots that showed up for their performance.  We decided to award
the robots anyways.  Unfortunately, not everyone aggreed with the decision.  Well, we slipped down that slippery slope,
and it is deffinately one area that needs to be addressed in the rules.  Hopefully next year, there will be more than 3
robots finishing the course, so that this problem will never come up again.

GPS coordinates will be given out again next year.  One idea that was being considered was not to provide GPS
coordinates to the bonus cones.

Pete



----- Original Message -----
From: "David P. Anderson" <dpa at io.isem.smu.edu>
To: "Ted Larson" <ted at larsonland.com>
Cc: <dprglist at dprg.org>
Sent: Tuesday, September 28, 2004 2:41 PM
Subject: Re: [DPRG] Re: Robo-magellan


> Hi Ted,
>
> Congratulations!  (Also very cool looking bot!)
>
> Thanks for the recap.
>
> So, it seems like GPS did not turn out to be too useful on this small of
> a scale?  That's interesting to know.  I think that I read that Doug
> is considering not even giving out GPS coordinates for next year's event,
> if I understood correctly.
>
> It does seem like the navigation contest, as it was orignally styled,
> became a cone-touching contest, (as I analyzed to no avail in the original
> request for comments, see SRS message #18121 and #18142).  And then, when
> nobody actually touched any cones, there was no objective method left to
> measure and rank the robots' actual navigation abilities, not surprisingly.
>
> Is than an accurate description?
>
> So at that point it became a subjective contest, with the judges having
> to make up unpublished criteria on the spot for a speculative conclusion
> as to who had the greatest chance of success:  "Most Likely to Succeed."
>
> Seems like a slipperly slope.
>
> Oh well, congratulations once again to you and Doug and Larry and to Servo
> Magazine and everyone who worked so hard to produce this event.  I think its
> a huge step in the direction of getting our experimental robots off their
> specialized contest courses and out into the real world.  Most cool.
>
> best,
> dpa
>
> Here's some links to the original observations on the Request for Comments,
> for folks not on the SRS email list:
>
> http://groups.yahoo.com/group/SeattleRobotics/message/18121
> http://groups.yahoo.com/group/SeattleRobotics/message/18142
>
>
> > It was an amazing event.  There were four cones laid out in the Seattle
> > center.  Three of them were bonus cones.  To win the $1000 prize, you had to
> > navigate to the final cone from the starting point faster than anyone else,
> > which was about 500 feet away, with many obstacles in the way...trees,
> > rocks, park benches, outdoor trash cans, decorative sculptures...a whole
> > bunch of stuff I never anticipated seeing.  There was definitely, no
> > straight-shot way to get from the start to the finish.  If you went to a
> > bonus cone first, you got a multiplier applied to your time.  The time was
> > scored on the best of best of three attempts.  Out of the 10 robots that
> > showed up, none of them touched a cone, on any attempt.   So, nobody won the
> > cash prize.   They ended up judging the winners of 1st, 2nd and 3rd place
> > based upon two criteria, who went the farthest, and who had the greatest
> > chance of success based upon their performance, and the technology equipped
> > on the robot.
> >
> > The event starting line was in a wooded area, with a narrow path leading out
> > of it, and up onto a lawn area.  Most of the robots never got out of the
> > woods by their 2nd attempt.  Most of them got hung up on a rock, or a tree
> > or hit a railing, or pole or something.  The trees and the big building near
> > the starting line, gave trouble to those who had a working GPS, so it made
> > for an interesting start.
> >
> > We had some major GPS problems on our robot (which is a long story) so, at
> > the last minute, I went in and modified our algorithm to navigate using only
> > the compass.  I totally anticipated we were going to loose, big time.
> > Anyway, it worked out better than expected, and we got very far, and took
> > 2nd place.  Needless to say, I was quite surprised.
> >
> > The guy who won the event got very close to the final cone, within 50 feet,
> > but the robot decided to take a path that lead him through a narrow space
> > between a tree and a garbage can, and he got stuck.  He didn't have much in
> > the way of proximity sensors on-board, but he did have a very-nice,
> > articulated frame design, that almost gave him the abilty to un-stick
> > himself.  I imagine next year he will have a sonar array on his bot.
> >
> > I think next year someone will win the cash prize for sure.  I thought the
> > event organizers did a great job for the first time of the event too.  There
> > were a lot of spectators....I would guess about 100 people or so.  It was
> > like watching a golf tournament.  This huge crowd following a robot through
> > the Seattle center to see how it would fare...and then you would hear a big
> > cheer when it cleared an obstacle, or ran into something....it was really
> > fun for the spectators as well as the competitors.  I think the SRS is onto
> > something good with this event.
> >
> > It was a lot of work to build the robot, and the competition was excellent!
> > Now that we have something built, we are already planning our strategy for
> > next year. Dave, I definitely would like to see you there next year with an
> > entry!
> >
> >
> > - Ted
> >
> > -----Original Message-----
> > From: David P. Anderson [mailto:dpa at io.isem.smu.edu]
> > Sent: Monday, September 27, 2004 12:11 PM
> > To: dprglist at dprg.org
> > Subject: [DPRG] Re: Robo-magellan
> >
> > Ted,
> >
> > So, how did you do this weekend?  Did you complete the course?
> > Did anyone?  Any post-event diagnoses?
> >
> > best,
> > dpa
> >
> >
> > >
> > >
> > > Thanks!  I guess I should take a photo or two with a coke can or
> > > something in there for reference.  The wheels are 12" in diameter.
> > > They are go-kart wheels.
> > >
> > >- Ted
> > >
> > -----Original Message-----
> > From: David P. Anderson [mailto:dpa at io.isem.smu.edu]
> > Sent: Wednesday, September 15, 2004 10:59 AM
> > To: ted at larsonland.com
> > Cc: dprglist at dprg.org
> > Subject: Re: [DPRG] Robo-magellan...
> >
> > Ted,
> >
> > Excellent!  It really doesn't look that heavy.  How large diameter are the
> > wheels?
> >
> > dpa
> >
> > > Dave,
> > >
> > > I finally put some pictures on my website of our Robo-Magellan robot.
> > > I will try to shoot some vide of it roaming around later in the week.
> > >
> > > http://www.tedlarson.com/robots/odyssey.htm
> > >
> > > Enjoy!
> > >
> > > - Ted
> > >
> > >
> > > -----Original Message-----
> > > From: David P. Anderson [mailto:dpa at io.isem.smu.edu]
> > > Sent: Wednesday, September 08, 2004 11:45 AM
> > > To: ted at larsonland.com
> > > Cc: dprglist at dprg.org
> > > Subject: Re: [DPRG] serial baudrate converter
> > >
> > > Ted, how's your Robomagellan comming?  Any pictures?
> > >
> > > thanks,
> > > dpa
> > >
> > _______________________________________________
> > DPRGlist mailing list
> > DPRGlist at dprg.org
> > http://list.dprg.org/mailman/listinfo/dprglist
> >
>
> _______________________________________________
> DPRGlist mailing list
> DPRGlist at dprg.org
> http://list.dprg.org/mailman/listinfo/dprglist
>
>


More information about the DPRG mailing list