AW: [robocup-small] Rule changes in SSL

[Raul Rojas]

There are two things here: a) the discussion about next year, b) ideas
about the future (which are sketchy right now).

My idea of an energy budget is not to limit engineering development.
It is to limit hardware wars. I also like to build faster robots, that
kick ten meters high. But the idea of the whole enterprise is to advance
multiagent robotics, to make robots collaborate and learn. Those are
the hard problems. Buying a 100 watt motor for the robot is much easier
than
making the robots plan and manage their energy (just like humans do).
The idea
of RoboCup is to build antropomorphic robots in the future, which will
exploit the strategies used by humans. 

Robotic competitions have always the problem that when someone finds a
shortcut that makes intelligence less important than motors, or kickers,
or whatever,
then the whole competition is much less interesting (and less relevant).
In 1999 someone put a
fast ventilator on a mid-size robot and blew the ball away from the
robot
towards the goal. It works. Terrific idea for winning the tournament -
the opponent
can do nothing. But that
is certainly a hardware-shortcut which makes intelligent behavior
unnecessary.
That's why ventilators were then banned in the mid-size league.

That is my motivation: eliminate or limit arms races. If there is a
better idea 
than an energy budget in order to achieve this, let us find it.

Raul Rojas
FU Fighters Team


-----Ursprüngliche Nachricht-----
Von: robocup-small-bounces at cc.gatech.edu
[mailto:robocup-small-bounces at cc.gatech.edu] Im Auftrag von Mike Licitra
Gesendet: Montag, 15. November 2004 00:53
An: robocup-small at cc.gatech.edu
Betreff: [robocup-small] Rule changes in SSL


We generally support (or at least accept) all the changes except for
point 7 (6 bots per team).

As some other teams mentioned, we also think that the field will become
a bit too crowded. More importantly, increasing the amount of robots
each year might give a disadvantage to financially limited teams. We are
trying hard to stretch our current budget to build 5 robots, there is
little chance for 6. We would have to use one of last years' robots as a
placeholder.
   
Maybe a temporary rule like "if one of the two competing teams has only
5 robots then the other team will also have to only use 5 robots" would
help. However, we realize that this might not be considered as fair for
all the teams which actually managed it to construct 6 robots.

Furthermore, we also believe that the suggested "energy consumption
limitation" for 2006 is partly a step in the wrong direction. For many
teams, RoboCup is just as much an electronic/mechanical engineering
challenge as it is on the software side. The development of faster
performing robots should be encouraged and we believe that they should
be used to their fullest physical potential and not artificially
limited. The goal should not be to make them more similar to the
physical performance of humans, but rather to outperform it.

We also don't think that these energy limitations really offer that much
of a new research potential since the simulation league has had similar
artifical limitations on their virtual physical constraints for many
years. Especially on the big field it is helpful to always run at
maximum performance in order to i.e. drive around the ball and stop it
from getting out of bounds. So please, let's leave artificial
limitations of this kind to simulation league.

Stefan Zickler, Mike Licitra

UB Robotics
University at Buffalo
http://www.eng.buffalo.edu/ubr/robocup.php