[robocup-nao] ball decision for 2010

[Mary-Anne Williams]

Hi Guys,

The variation is not just in weight (with or without gel) but texture. Some
are shiny and others flat - with dramatic implications for vision systems.

An advantage of the tennis ball in the Humanoid KidSize League is that it
has a flat surface and easier to see.

Franklin have several versions.

*Flat Franklins* http://www.streethockeyking.com/frnhlhidesth1.html

*Shiny Franklins* http://streethockeyking.com/fragprohides.html

Cheerio,

Mary-Anne.

-----------
Professor Mary-Anne Williams
Gates 222, Stanford Artificial Intelligence Lab
Stanford University, CA 94305-8610
Telephone: (650) 391-4774
Google Voice:  (650) 646-4391
eFax:  (636) 246-5226
http://innovation.it.uts.edu.au/Mary-Anne



2009/12/14 Nao-Team HTWK <naohtwk at googlemail.com>

> Hi All,
>
>
>
> The problem with the hockey balls is, that there seem to be two hugely
> different versions: “normal” balls and balls with some anti-bounce gel
> system (often called AGS, active gravity system).
>
> The Mylec ball from the first link Tekin sent around is a no-bounce (i.e.
> gel filled) ball, the A&R ball from the second link doesn’t seem to be a gel
> ball.
>
> I guess, as you pointed out the better back-bouncing properties of the
> hockey ball, you tested a gel-filled ball.
>
> Gel-filled balls are available in Germany mainly from Franklin, however,
> they provide different densities: super high density (105g, used in roller
> hockey), high density and medium high density (45g).
>
> I guess the “correct” ball would be the high density ball whose exact
> weight I couldn’t find yet.
>
>
>
> Rico
>
>
>
>
>
> *From:* robocup-nao-bounces at cc.gatech.edu [mailto:
> robocup-nao-bounces at cc.gatech.edu] *On Behalf Of *Tekin Mericli
> *Sent:* Samstag, 12. Dezember 2009 13:02
> *To:* Oskar von Stryk
> *Cc:* Humanoid League Mailing List; RoboCup Nao Mailing List
> *Subject:* Re: [robocup-nao] ball decision for 2010
>
>
>
> Hi Oskar,
>
> On Sat, Dec 12, 2009 at 1:11 PM, Oskar von Stryk <
> stryk at sim.tu-darmstadt.de> wrote:
>
> Hi Tekin,
>
> As you probably know, the RoboCup Humanoid KidSize League uses
> an orange standard tennis ball.
>
> To my (limited) knowledge, the standard data for tennis balls are
> - diameter: between 6.35 und 6.67 cm
> - weight: between 56.7 und 58.5 g
>
> The data you mention for an orange street hockey ball
> (65mm in diameter and 55g weight) matches the
> above data of an orange tennis ball pretty well.
>
>
> You are right, the dimensions of a street hockey ball are very similar to
> that of a tennis ball. Let me try to answer your questions below.
>
>
>
> Therefore I was wondering about the following:
>
> 1) Why did you select a hockey and not a tennis ball
>        as it is used in the Humanoid KidSize League?
>
>
> We had some discussion about whether switching to a tennis ball after the
> original ball would be a good idea. You are right, tennis balls are common;
> however, they have a lot of variance in how they are made (fuzziness,
> weight, color, etc.).  Thomas pointed out that the orange tennis balls are
> relatively tough to get - I think they are only available from a dealer in
> Canada. On the other hand, the street hockey balls are available (almost)
> worldwide and they have very similar texture characteristics.
>
>
>
> 2) Did you compare the performance of the hockey ball
>        with a tennis ball on a RoboCup soccer field?
>        And, if so, what were the results?
>
>        (Besides size and weight, tennis balls must satisfy certain
>        back-bouncing abilities from the ground and will probably
>        behave differently than hockey balls in robot soccer.)
>
>
> Yes, the performances of various balls including our old ball, tennis ball,
> street hockey ball, and juggling ball were compared in terms of visibility
> from longer distances, rolling straight, rolling distance, and amount of
> back-bouncing. Since the street hockey ball can be seen fairly well from
> longer distances, rolls straighter compared to our old ball, does not bounce
> that much so that incoming ball control would be easier, and being easily
> available worldwide, we decided to proceed with the street hockey ball for
> now.
>
>
>
> 3) If teams order a standard orange street hockey ball from
>        a store which does not give the precise diameter
>        and weight of the ball, which range of variation
>        in size and weight must be expected?
>
>        From a quick internet search it seems that there
>        is more variation in street hockey balls than
>        in tennis balls. I found: diameter 6.35-7cm, weight: up to 85g.
>
>        A weight difference between 55g and 85g can make
>        an enormous difference to the robot performance (e.g.
>        when kicking the ball)!
>
>        In the Standard Platform League where only software should count
>        I would expect also a standard ball that teams can, e.g.,
>        remotely participate in local RoboCup Open competitions.
>        With a large variation in ball weight, this will be difficult.
>
>
> We pointed to some certain brands, so the balls you get won't have any
> problems if you specifically look for those brands or if you find another
> brand and ask the store about the dimensions to be sure before you purchase
> them. One such brand is Mylec:
>
> http://www.mylec.com/mylechockey/Products/tabid/102/Default.aspx
>
>
>
> Thanks for your time and efforts.
>
>
> You are very welcome.
>
>
> Cheers,
>
> Tekin
>
> _______________________________________________
> robocup-nao mailing list
> robocup-nao at cc.gatech.edu
> https://lists.cc.gatech.edu/mailman/listinfo/robocup-nao
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.cc.gatech.edu/pipermail/robocup-nao/attachments/20091228/d13e72ef/attachment.html