Loading ...
Sorry, an error occurred while loading the content.

121Re: [sfrsa] Annual Robot Games

Expand Messages
  • Paul F. Grayson
    Jul 30, 2000
    • 0 Attachment
      One way to structure contest rules is to give points for each of the
      tasks... if a robot can not do a particular task... is misses those points,
      but can still score with the things it can do.

      Paul F. Grayson - Chief Engineer
      AMERICAN INDUSTRIAL MAGIC
      1892 Pinewood Ave.
      Traverse City, MI 49684-9022
      (231) 946-0187, FAX (231) 946-1122
      pgrayson@...
      ----- Original Message -----
      From: Jeanette Eya-Zeissig <jzeissig@...>
      To: <sfrsa@egroups.com>
      Sent: Monday, July 31, 2000 1:53 AM
      Subject: Re: [sfrsa] Annual Robot Games


      > For my proposed event at the Robot Games, I think I would like to take on
      > all comers that don't compete in the Sumo, Lego, or other win-lose events.
      > There has been mention of some of the concerns that I brought up in my
      > previous post on the Seattle Robot Society discussion group. I hadn't
      been
      > aware of these when I did the post.
      >
      > The S.R.S. list of events for Robothon 2001 shows a floor exercise event
      > which is just a 10' square enclosed arena with a smooth floor.
      > Participants can launch their 'bots for 5 min. sessions to show what they
      > can do. The emphasis is on autonomous robots, but they are not
      restricting
      > participation to that category. This would certainly be easy to
      implement.
      >
      > I think we could spice this up by providing some portable/adjustable
      > accessories that participants could use to customize the arena. What I
      > have in mind is:
      >
      > a) an adjustable-slope ramp to test climbing ability and traction
      >
      > b) a section of carpet or astroturf
      >
      > c) an adjustable step height staircase
      >
      > d) a gravel/rock/or sand pan
      >
      > e) a shallow water hazard/pan
      >
      > f) a set of moveable block/post obstacles
      >
      > g) a section of corrugated roofing panel
      >
      > h) anything else that a participant wants to lug to the arena as
      > long as
      > they are willing to take it away again upon request. (Tall grass is
      > a wonderfully amusing challenge for my walker, but I can't think of
      > any way to incorporate it in this context.)
      >
      > Most of these are things that I've been meaning to make anyway to
      calibrate
      > 'bot performance when making mechanical or software changes, so it would
      be
      > a good opportunity to quit procrastinating. Some, like the corrugated
      > panel and carpeting, are just junk that I have laying around.
      >
      > The point of this event is to maximize participation. I've gone back and
      > looked over the postings on last year's robot games. I think Cliff
      pointed
      > out that the obstacle course looked hokey because people were throwing
      > shoes and socks in the arena. Stan didn't participate because the rules
      > weren't posted clearly and well in advance. These are points well taken.
      >
      > Although my 'bot was fully functional, and I had planned to attend, I
      > didn't; because when the rules were finally posted, the stair climbing
      > portion of the obstacle course seemed to call for a 2" leg lift, and my
      > 'bot couldn't make it without hours of rework. I figured it would be a
      > DNF. That was a mistake: I should have participated anyway. The point
      is
      > that if you make the rules too precise you exclude participants; hence an
      > adjustable stair climb that can accommodate the legginess of the 'bot. Or
      > no stair climb at all if the 'bot doesn't do stair climbs. I would rather
      > err a little bit on the side of hokiness if we could attract a participant
      > with a 'bot that could wander the arena using dead-reckoning and wind up
      > anywhere near its starting point. On the other hand, we would have some
      > toys handy so the shoes and socks wouldn't be necessary if somebody wanted
      > to show off their 'bot's obstacle avoidance prowess.
      >
      > My biggest worry in all of this is how to ensure that the maximum
      > participation is achieved while simultaneously making sure that all the
      > 'bots get a satisfying workout. This is where I'll really need some help
      > and feedback. Is it unrealistic to attempt a modifiable arena if this is
      > only a one-day event? How many participants should be anticipated?
      Should
      > we try to get an advance estimate by e-mail polling or pre-registration?
      > Auto clubs are doing this now for inter and intra-club race events. Are
      > all the events running sequentially or concurrently? So let's hear from
      > those of you with some experience. I've only attended one robot event:
      > the Japanese Sumo Challenge. This was a straightforward elimination event
      > that only took a couple of hours, including some pre-demonstrations and
      > some exciting 'bot free-for-alls and 'bot-human shoving matches after the
      > main event.
      >
      > Finally, what is the reason I want to do this event? I've been working on
      > my robot for years. I started out with only CMOS logic gates for control
      > and actually achieved some things that could only be accomplished by
      > somebody who didn't know any better. I've had to completely start over
      > when my surplus supplier of linear stepper-motors went out of business:
      > new mechanical design, new electronics, the works. I think I started the
      > whole thing in 1993. Anyway, in all that time, I never saw a real, honest
      > 'bot in the round until I went to a Home-Brew Robotics Society meeting
      > sometime in '98. What I saw there was about as far along as what I was
      > working on at the time. I've seen a few more at S.F.R.S.A. meetings. I
      > saw maybe two dozen 'bots at the Sumo event, and they resembled my 'bot
      > about as much as a bulldozer resembles a cockroach. I was really excited
      > that the Alan Alda TV special was showing a "robo-roach," and was
      > incredulous when it turned out that the damn thing couldn't even stand up,
      > let alone walk. My 'bot could run circles around it. If you listen to
      the
      > news media, you would think that robots are as common as pigeon droppings,
      > but it's not so. Functioning, autonomous robots are very, very, very thin
      > on the ground. My wife and I joke about the family 'bot and the
      > neighborhood 'bot; but really, we don't know any other family or
      > neighborhood that has a 'bot, and I suspect ours is the only one in town.
      > If you have one, you are about 3+ standard deviations out there on some
      > scale that psychometrics has yet to identify. But I know those 'bots are
      > lurking somewhere. Maybe they're all up in Seattle, I don't know. What I
      > do know is that I want to see them in action, and it looks like the
      easiest
      > way to do that is to host this event.
      >
      > John Zeissig
      >
      > <http://home.att.net/~jZeissig>
      >
      > <mailto:jZeissig@...>
      >
      >
      > To see an older version of the 'bot
      <http://home.att.net/~jZeissig/Bot.html>
      >
      >
      >
      >
      >
      > Be sure to visit our web site at http://www.robots.org
      >
      >
    • Show all 11 messages in this topic