Subject:
|
Re: Soccer-Playing Robots (Question for Mario Ferrari & Others)
|
Newsgroups:
|
lugnet.robotics
|
Date:
|
Wed, 30 Aug 2000 22:18:35 GMT
|
Viewed:
|
829 times
|
| |
|
|
In lugnet.robotics, Mario Ferrari writes:
> OneTimeCRX <lego-robotics@crynwr.com> wrote:
> > I think the person you're looking for is Henrik Lund, with his website • here:
> > http://www.daimi.au.dk/~hhl/MindFest/
> > Those soccer robots were impressive!
>
> You're right, the soccer-playing robots were from Henrik Lund & Luigi
> Pagliarini, and they were really impressive.
>
> Anyway, a few months ago Marco Berti and Giulio Ferrari made their own
> version of the game to be shown at an exhibit, and worked great.
>
> Marco built the bots, the field and the IR ball, and Giulio programmed them
> using NQC. If you need more info you can reach them at marco.berti@itlug.org
> and giulio.ferrari@itlug.org.
>
> HTH
>
> Ciao
> Mario
As Mario and Giulio made my name I have been forced to beat my laziness and
upload some picture of the IR emiiting ball (sorry it has not been made with
Lego parts!) and of the robots.
You may find these picture here
http://www.brickshelf.com/cgi-bin/gallery.cgi?f=1223
Honestly Giulio did much more than simply helping me with NQC: we spent one
night togheter building and testing the hardware and the software.
The robots have two IR sensors: the one on the front seek for the ball and
detetect when the ball is catch, the one in the middle looking down read the
color of the playing field and detect the direction to go (the "black" bot goes
towards the clear part of the field and the "white" one towards the black).
The playing field has been made with an ink jet plotter and has a gradient from
80% black to 10% black.
The real problem was to have enough "IR light" from the ball.
I solved the problem "squeezing" 28 IR led , 6 high bright led , 4 x 350 mAh
sanyo nicd cells , a recharge plug and a switch inside a 5 cm (2 inches)
transparent ball (stolen to my son Carlo).
This give enough light power for the IR sensor to detect the ball up to 3
meters in a normal room light and a working time of about one hour and half
when you have fully charged the nicd.
The playing field layed on a special shaped basement that has 2 degrees of
slope toward the centre from all the four side, so if the ball is "lost" it
will return to the centre.
The two bumpers use two cybermaster switch sensor that are both connected to
the third input : as they have different resistance value the software knows if
have been pressed the right or the left bumper
They are far from being perfect ... but are a lot of fun to be looked at.
The project was started to display the potemtial of the Lego based robots in a
"non Lego maniac environment" and I have been quite happy as they rised a lot
of interest .
Sooner or later we will improve this project ... maybe!
Marco
|
|
Message is in Reply To:
5 Messages in This Thread:
   
- Entire Thread on One Page:
- Nested:
All | Brief | Compact | Dots
Linear:
All | Brief | Compact
|
|
|
|
