[robocup-legged] 11x11 RoboCup match
William Uther
willu.mailingLists at cse.unsw.edu.au
Fri Jan 6 19:55:00 EST 2006
Hi all,
I'm glad there is some interest in the 11x11 match. Issues that
have been raised:
- Lighting
Yes, we will need enough lighting for the AIBOs to see.
- Field
The 2005 mid-sized field was about double the size of the legged
league field (linear dimensions). I suggest we just go with that.
The only change would be the size of the centre circle. The mid
sized one is a little large for AIBOs (2m diameter as opposed to 30cm
diameter). I suggest say, 60cm diameter?
We could use mid-sized locations for the 'restart points', but use
them as the end-points of the legged-league 'throw-in lines'.
- Landmarks
My original thought was that we would have Mid-Sized landmarks
unmodified, but I have no objections to making mid-size-league sized
landmarks, but with legged-league style markings. We could go back
to having 6 landmarks as well. I think that the increased landmark
size should enough that these can be seen across the field, but we
need to check (I haven't head from anyone with a mid-sized landmark
and an AIBO yet). We've doubled the linear dimensions of the field,
so landmarks ~4 times higher and double the diameter should still be
visible, right? Yes, there will be an associated loss of absolute
localisation accuracy, but the relative accuracy should be similar.
- Ball
I was assuming we stay with a legged league ball. You will not be
able to see it from very far away, but you will be able to spread
your team out and communicate to find it. We don't want to have to
re-develop ball manipulation skills.
- Goals
I have no idea what width to make the goals. The goalie is still
an AIBO with limited speed, but a goal the size of the legged league
goal would look silly. Making a third size of goal somewhere in
between means we make to make another set of goals. I'd start with
mid-size goals, but am open to suggestions.
- Communications
I was planning to have something simple like that used in the
rUNSWift/NUBots open challenge last year. That was a UDP broadcast
packet. Each robot would broadcast:
- Its team and player number
- Its own location and covariance matrix
- Whether the robot can sense the ball, and if so, the ball's mean
location and covariance matrix
- "sensing the ball" includes both visual sensing and whatever
other mechanisms you have to detect the ball, e.g. the mouth or IR
sensor during a grab.
- The current destination of this robot. This might be the
location of the ball, or a location on the field.
I don't know that we need much else. We might want an: "I'm going to
kick the ball here", section in the packet. I wouldn't want too much
more than that: Keep It Simple and Stupid. We didn't used any more
than this in our pick-up open challenge last year. Teams that want
to use more certainly could - amongst those robots using that code base.
- Game Controller
Yes, we'd need a modified game controller that could handle
11x11. I don't think the packet format needs to be changed, except
for increasing the size of one array. We could come up with a
general format that included the number of players on each team.
Then the same packet would be used in normal games and the 11x11
game. The tricker part is re-designing the UI in the GameController
program itself, but I don't see this as a huge issue.
- Rules
Rules would be similar to the legged league rules. i.e. there
would be no off-side concept. The only ones that I think really need
to change are kick-off rules. Even those changes can be kept simple:
rather than being behind the 1/4 line, you must be 2m from the centre
line.
- Physical setup.
This is the part that requires extra work. And the answer at this
point is: I don't know who would build the landmarks, make numbers,
etc. If the landmarks are simply re-coloured mid-sized landmarks
then that isn't much extra work - it is really extra cost, and that
decision is made by people on other committees.
Finally, I'd like to comment on expected performance. Simulation
league teams have very high standards of play for 11x11 matches. We
have high quality low level skills. The real issue is testing time,
especially given the pickup part of the demonstration.
Because of the almost complete lack of testing time, I'm not going to
be disappointed if the standard isn't wonderful. While we want to
make this as good as we can, if we worry too much about all the
things that could go wrong then people might decide not to bother. I
don't want that to happen. It is much better to have something that
we can build on for future years than to make this a perfect game of
soccer.
Is there anything I missed?
Will :-}
More information about the robocup-legged
mailing list