[robocup-small] Looking ahead to RoboCup 2006 - vienna cubes

[Walter Scherer]

Hi all!

First of all we would like to apologize for submitting our official feedback a
few days past the original deadline. There was a lot of discussion going on
within our team and we wanted to give all our team members enough time to
participate. Some team members might have individual feedback to give, they will
submit it as soon as possible.

Feedback for RoboCup 2005:
We think that RoboCup 2005 was very well organized. The field illumination and
construction was perfect but we (and some other teams, we guess) could have used
a little more space to work around the field. Maybe next year we could have more
space between two fields and between team tables and the field as well because
sometimes you could hardly move. Referees had a hard time not to become
entangled in the cables and to have a free passage around the field. We wonder
if the organisation committee can grant us space to set up a test field where
new developments (like local vision robots) can be tested.

We would appreciate longer opening hours of the venue, if that is possible.

Here are the vienna cubes' comments on the rules:

1) Walls and Field Setup:
summary of our 2 possibilities:
a) small 90° walls, no out-of-game areas, field size as is
b) no walls, increase field size to 1/2 mid-size field

Maybe the league is not yet ready for removing walls altogether. So our first
proposal for next year is to build small 90° walls that would deflect light
passes or shots and call the area behind the walls as out-of-game. During last
RoboCup our team played a kind of test match against Big Red where we did not
care for the field boundaries and just let the ball run. We must say that it was
among the best, most interesting and funniest games we have ever seen. But on
the other hand the league and its teams should focus on developing really
appealing tactics and etablishing good teamplay. The good thing about having no
walls is that your team is forced to play accurately (or at least: shoot
accurately). Still, we think that only direct shots should be counted as goals
(direct in this case means no deflections off walls) to keep a certain
importance of team play.

As another possibility we would also agree to enlarge the field to approx. one
half of the mid-size field. This would mean no walls are necessary, but teams
are forced to play well together to win games. Increasing the field size to
mid-size field would be too risky. We should make the field bigger gradually
year by year, if we go for the "bigger-is-better option".

2) Dribbler:
summary:
- active and passive dribblers allowed
- dribbling restriction (50 cm) for all kinds of dribblers

A well-constructed dribbler allows you to receive passes and keep the ball which
is a prerequisite for team play. That's why we think we should keep the dribbler
and its restrictions and rules as they are at the moment. The 50 cm dribbling
restriction is a good thing to have and although it is hard for the referee to
judge, we think it is just fair that we keep it the same for all kinds of
dribbling devices.

3) Speed of robots (and its possible limitation):
summary:
- no speed limitations
- speed needs control

Because a robot just does not get tired like humans do, why would we like to
limit the robots' energy budget and/or their speed by an external
vision-referee? Robots have advantages we have to use, but we have to use them
in a clever manner. That means if a team goes for e.g. speed it has to be sure
that it can control this speed. If the robots go fast and play well together,
well, then they will probably win and there is nothing wrong with that. The peak
of speed is now more or less reached, now we must concentrate on controlling the
speed to make it a real advantage for the team and consequently for the league.

4) Restart position:
summary:
- change to always start 10 cm away from the boundary lines for all restarts

We agree with Prof. BengKiat's idea to change the rules accordingly.

5) Official colours:
summary:
- issue official colours for team colours and ball

We support this year's official colours for the teams and the ball. Teams have
to change their team blobs to the official colour unless both teams agree on
keeping their colour. All other colours should be chosen freely by the team (in
accordance with the rules of course). That means pink is all right with us.

6) Mandatory check of robots:
summary:
- check before each game

Checking robots before each game costs much time but we think that it will stop
unnecessary discussions afterwards. Especially if the checks are carried out by
officials. This should be done before every game (as demonstrated in the finals
in Osaka).

7) Video evidence:
summary:
- no official video evidence
- what the referee says goes

We think an official video would only lead to new discussions and waste precious
game time. So basically: More problems than it solves. We could introduce a
substitution referee (like in real soccer, the one who announces over time with
a display). He should be the communication interface between the teams and the
main referee so that the main referee can fully concentrate on the game. If a
critical situation occurs the main referee can fall back at the substitution
referee to see what she/he thinks and then decide.

8) Penalty kick:
keep as is

9) Chip kick:
summary:
- always allowed

Chip kick should always be allowed (to score a goal, too, unless it is a direct
goal from an indirect restart). We think that the chip kick very much adds to
the excitement of the game. This development should not be undone.

10) Dangerous robots:
During some games robots of some teams were losing parts quite frequently and
shot them into the audience. This ought to be stopped. If a robot loses parts
the game should be halted, the robot should be sent off for repair immediately
and should be allowed to join the game later when it has been repaired (similar
to an slight injury in real soccer). We are clearly against repairing robots on
the field because you just have the time to fix a problem with a piece of tape
and that is very dangerous anyway. Teams that have a few substitution players
should be allowed to change the "injured" robot for a good one. If robots lose
parts frequently the team should be booked (or a robot should be sent off with a
red card in extreme cases) because of dangerous robots. This rule was enforced
anyways in this year's competetion as far as we know but maybe not forcefully
enough.

11) Local vision:
summary:
- local vision cannot yet compete with global vision systems
- lots of rule amendments to be made
- => gradually introduce local vision (maybe in some test games/challenges first)

Currently we consider local vision as no option for the SSL. This is because the
league has to overcome difficulties in team play (see David M. Chelberg's
comments) before we can switch to local vision. Besides, local vision will
require quite a few rule changes. We are not sure how much experience other
teams have using local vision in the SSL but we guess global vision was always
considered superior to local because of the current (rather small) field size.
But as field size grows we believe that someday we will end up in local vision.
Therefore we support the idea of gradually introducing local vision on test
fields first and later on incorporate local vision into the games.

12) Challenges:
summary:
- state clearly where they take place and when
- challenges as field test for new developments
- local vision challenge?

We generally support the idea of challenges (for they are part of research)
although we can also understand some teams don't have the capacities to take
part. We would like the challenges to be a sign for teams in which direction
research in the SSL currently goes. If we want to introduce local vision, it
might be a good idea to start it as a challenge event and see what happens.

13) Kick-Off:
summary:
- allow executing robot to cross the middle line for kick-off

Generally speaking the kick-off rule is all right. Unfortunately, most teams
just showed two options to carry it out. One is directly shoot into the
opponents field, the other one is try to pass to the touch line and hope that
another robot will catch the ball before it goes out of bounds. Both options are
very aggressive. We would like to have the possibility to establish good pass
game from kick-off (cp. real soccer) which implies passes into the own half.
Therefore we would need to cross the middle line to get around the ball.

14) Energy budget:
summary:
- no limitations
- battery changes during half time or timeout allowed

We think that limitations concerning batteries would be a step into the wrong
direction and that the limits for size of robots are enough. We believe that all
teams enjoyed the freedom to choose whatever affordable equipment they wanted as
long as it fitted into the robot. In the past this has led to astonishing
developments and we strongly agree to stay on this track.

15) Constuction of goals:
summary:
- crossbar and posts should be made of solid, round material

Until now the referee could barely judge if she/he should award a goal when the
ball was chipped into the goal but touched the upper net. A solid round crossbar
(made of wood or aluminium) would solve this problem. Maybe we could also use
round material for posts to make it look even more like soccer.

Regards,

Walter Scherer
vienna cubes