Subject:
|
Re: Selflearning Scout robot
|
Newsgroups:
|
lugnet.robotics
|
Date:
|
Tue, 23 Nov 1999 21:20:42 GMT
|
Original-From:
|
Bert van Dam <bvandam@westbrabant.!ihatespam!net>
|
Viewed:
|
471 times
|
| |
|
|
Hi Mario,
Yes, you did prove me wrong, and I promised to put a link to your robot on
my site. But then my harddrive crashed and I lost all my mail, including
your address. But I'll live up to my promise and add you right now!
This particular braintype can be used for a bit more complex behaviour, but
that's rather limited. I'd opt for a neural brain for that but I don't think
I can fit one in the 1k memory of the Scout :)
CU Bert
Lego robots and artificial intelligence www.bvandam.net
-----Oorspronkelijk bericht-----
Van: Mario Ferrari <mario.ferrari@edis.it>
Aan: lego-robotics@crynwr.com <lego-robotics@crynwr.com>
Datum: dinsdag 23 november 1999 9:46
Onderwerp: Re: Selflearning Scout robot
> Bert van Dam <bvandam@concepts.nl> wrote:
> > Using the new Lego SDK for the Scout I've built and programmed a self
> > learning Scout robot. The artificial intelligence of the Scout is based • on
> a
> > roulette wheel brain. The small robot will drive around and learn how to
> > avoid colliding into walls (objects).
> >
> > Both the robot itself (ai-wall avoider) as well as the theory of roulette
> > wheel brains (ai-roulette wheel) can be seen on my website.
>
> Hi Bert,
>
> You did already dig deeply into the Scout!
> Very interesting. It surely applies to more complex behaviour too.
> I used a similar approach with my learning brick sorter, but I never • thought
> to the roulette wheel, it's a very clear metaphore.
>
> Mario
> (the one who proved you wrong about the possibility of doing a Lego robot
> that shifts its weight to walk with just too legs :-) )
>
> Web page: http://www.geocities.com/~marioferrari
> LUGNET member page: http://www.lugnet.com/people/members/?m=22
|
|
Message has 1 Reply:
 | | Re: Selflearning Scout robot
|
| (...) Thanks! Excuse me, I couldn't resist poking you about that :-) (...) think (...) Did you succeed in implementing a neural net on the RCX? I'm sure it IS possible, if not with NQC surely with legOS (though with NQC too we succeeded in emulating (...) (25 years ago, 24-Nov-99, to lugnet.robotics)
|
3 Messages in This Thread:
 
- Entire Thread on One Page:
- Nested:
All | Brief | Compact | Dots
Linear:
All | Brief | Compact
|
|
|
|
