[robocup-small] Rule changes for 2005

[Frank Sill]

Hi Robocuplers,

After reading the proposed rule changes and all these interesting 
comments, we want to add our point of view.

First, we don't understand why the team size should be changed to six or 
more robots. There is no advantage in having more robots. All the 
strategies work also with five robots. Having six robots would only 
increase the costs for smaller teams.

One of the main problems of the current RoboCup rules is that the ball 
leaves the field too often. Some teams suggested a bigger field and 
walls to keep the ball in game. This would solve the problem but not the 
reasons for it. In our opinion, teams should improve their strategy to 
keep the ball in game. A bigger field has already been introduced one 
year ago.  The teams did not have the time to change their strategies 
yet. Additionally, it is much harder to find a place for such a big 
field. Even middle size league teams often use only half of the field 
size for training.
We propose the following solution:
bringing the ball out should be a bigger disadvantage than it is now!
After the ball is out, the kick in team can have all robots within 40 cm 
range around the ball the other team can only have two robots between 40 
and 80 cm around the ball. The other robots can be placed somewhere on 
the field.
This solves three problems:
- Penalty: Letting the ball go out is a penalty now
- Room: We gain more room between the robots on kick in - no more 
circles of robots!
- Game flow: The ball can not bounce against another robot and go out again.

We don't see a big problem in using horizontal dribbling devices, 
provided, that another robot is still able to catch the ball from 
another robot. Therefore, the dribbling device should not cover any part 
of the ball. This means, that it has to be places at the horizontal axis 
of the ball.
Vertical dribblers give a very big advantage over other robots. It is 
nearly impossible to take a ball away from such a robot because it can 
maneuver at the same speed as without carrying a ball.


The idea behind the energy budget solution of Raul is not bad. But the 
realization over an automatic referee is very problematic:

1) The referee has to distinguish every robot. Today this is done over the 
marks on the robot. But what happens, if all robots in a team have the same 
marks (we tested that to disburden our vision) ?
2) How it could be assured, that the referee always get the right values? For 
example, if the robots accelerate they don’t have an equal velocity. How the 
vision can distinguish between pass, hard kicks and chip kicks (robot in the 
way of the moved ball) ? What happens, if the tracking of the referee vision 
looses the robot?
3) If you punish sprints it could be possible, that no robot has interests, to 
save the ball before going out.

The question is why do we need such a complicated solution? You shouldn’t 
forget that we have an energy budget. Everybody has to work with the given 
size and the weight of great batteries. Nobody can choose i.e. a car battery. 
An easier solution for the energy budget could it be, if we limit the capacity 
of the used batteries. This limitation should be coupled with the weight of 
the robots (as some solutions mentioned in some mail before)


Last but not least, we see no reason why newer RF technologies, like 
bluetooth, should not be allowed in RoboCup. The RoboCup league was 
founded to test and use new technologies. And see how they work. 
Bluetooth does not interfere with other RF technologies. The used 
transmission power is to small. It should be up to the team to deal with 
the weaknesses of the used RF technology.
In our opinion, all technologies that comply to public RF standards 
should be allowed. This includes WLAN, DECT and Bluetooth.

In general, team specific problems should not be solved with new rules. 
The cause should be solved, not the effect.


Cool Running
University of Rostock, Germany
http://robocup.e-technik.uni-rostock.de