Subject:
|
Re: robotica on TLC
|
Newsgroups:
|
lugnet.robotics
|
Date:
|
Fri, 6 Apr 2001 19:22:40 GMT
|
Reply-To:
|
micahx@kih.netNOMORESPAM
|
Viewed:
|
4283 times
|
| |
|
|
Steve Baker wrote:
>
> "Micah J. Mabelitini" wrote:
> >
> > Steve Baker wrote:
> > >
> > > What I think it would take would be to add a rule that every robot must
> > > contain a GPS unit and broadcast it's position continually for the other
> > > competitor to pick up.
> >
> > GPS wouldn't work. Even though the government disabled Selective
> > Availability last year, commercial GPS still doesn't achieve an accuracy
> > better than several meters.
>
> Differential GPS (where you add a 'ground station') gives you plenty
> of precision - even without the low order bit scrambling thing.
>
> It's not good for telling you where you are relative to the planet - but
> relative to the ground station, it gets you within a couple of centimeters...
> plenty for these meter-wide 'robots'.
Yeah, DGPS would probably be accurate enough, assuming you could get
around the no-lock problem.
> I would imagine that for GPS to be suitable
> > in such a competition, it would require an accuracy of no more than a
> > few inches, and would require that all of the platforms were equipped
> > with *exactly* identical GPS modules manufactured with very rigorous
> > tolerances.
>
> The TV company would presumably supply those in order to avoid cheating...
> and they could be pretty high-end units.
>
> > Also, getting a lock inside any building is difficult and
> > unreliable at best, and it appears that these types of competitions are
> > generally held in large steel buildings, an environment totally
> > unsuitable for receiving a GPS lock.
>
> Dunno about that...you could well be right. Does the presence of the
> ground station inside the building fix that? I have no idea.
I believe the on-board GPS receiver would still need to fix on the
Navstar constellation, regardless of whether or not it was a
DGPS-augmented system. That's not going to happen in a large
warehouse-type building.
> > What about some sort of standardized IR beacon pulse on an
> > omnidirectional emitter, and equipping the platforms with a servo-driven
> > rotating unidirectional detector? Since all of these competitions seem
> > to be one vs. one
>
> (No - BattleBots does an 8 robot 'rumble' from time to time - and the
> British Robot Wars series has several 'games' where there are many
> machines in the arena at once - with a set of robots designed by
> the BBC (The "house robots") in nearly every competition.)
I wasn't aware of the 8 robot thing (I've only seen BattleBots
once)...that sort of situation would definitely complicate matters.
> > have standard emitter designs with two pulse
> > frequency settings. Then, one competitor could set their locator system
> > to Mode 1, and the other to Mode 2, preventing a platform's emitter from
> > interfering with its own detector.
>
> Yes - something like that would work - but anything on the outside of the
> case would get smashed off in short order I think...also many of the
> designs have to work upside-down as well as right side up.
Yeah, I considered that. It would no doubt be hard to engineer such a
system in a rugged enough fashion to be reasonably reliable.
> Since the arena is fixed, you could put light detectors in the floor
> and have the TV company track where the robots are by knowing which
> detectors are covered up...then broadcasting the locations of the
> two machines by radio...there are LOTS of ways to make that kind
> of thing work.
>
> Anyway - the point is that to make a viable "true" robotic competition,
> there would have to be some kind of mandatory tracking mechanism...you
> could argue indefinitely about what it should be - but it's pretty
> clear you could solve it.
Agreed. I think you're on the right track with the whole "DGPS stations
inside the building" thing though. But, since you dont need absolute
global positioning, only a relative position on a local coordinate grid,
perhaps eliminate the GPS altogether and just set up triangulation
beacons around the arena?
--
Regards
Micah J. Mabelitini - LUGNET #918
The University of Kentucky
SECC Middlesboro Academic Skills Resource Center
accutron@kih.net - http://www.users.kih.net/~micahx/
|
|
Message is in Reply To:
 | | Re: robotica on TLC
|
| (...) Differential GPS (where you add a 'ground station') gives you plenty of precision - even without the low order bit scrambling thing. It's not good for telling you where you are relative to the planet - but relative to the ground station, it (...) (24 years ago, 6-Apr-01, to lugnet.robotics)
|
38 Messages in This Thread:
 
 
- Entire Thread on One Page:
- Nested:
All | Brief | Compact | Dots
Linear:
All | Brief | Compact
|
|
|
|
Active threads in Robotics
|
|
|
|
|
|
