Bowling machine no match for humankind
Nice try, there, "Harry" (if that is your real name), but don't quit your day job yet. In the first ever contest to
pit human versus machine in heated bowling combat, Danny Wiseman of Baltimore solidly trounced "Harry" the ball-thrower
at the United States Bowling Congress Research and Testing Facility (do our tax dollars pay for these things?). In a
race to five strikes, Wiseman was five for five while poor Harry only struck once out of its five attempts. Developed
by USBC engineers primarily for testing new equipment, the bowling machine had its chance to shine during its 15
minutes of fame… and blew it. At least it's good to know that we can take our robotic overlords at something,
anyway.
[Via The Raw Feed]

















Here are two links to the Throbot, another bowling robot, used in the R&D department of the major bowling equipment company Brunswick:
http://www.memagazine.org/backissues/january99/departments/news_notes/news_note.html
http://www.findarticles.com/p/articles/mi_m1511/is_3_21/ai_59535395
"The United States Bowling Congress officially launched Jan. 1, 2005, as the organization to serve amateur adult and youth bowlers in the United States. It resulted from the merger of the American Bowling Congress, Women's International Bowling Congress, Young American Bowling Alliance and USA Bowling. Today, USBC serves more than three million members.
USBC is the national governing body for bowling as recognized by the United States Olympic Committee.
USBC will stand for values that include: credibility, dedication, excellence, heritage, inclusiveness, integrity, philanthropy and sportsmanship."
http://www.bowl.com/aboutUSBC/main.aspx
"the bowling machine had its chance to shine during its 15 minutes of fame…"
Should have said '15 minutes of frame'...yeah...I know....that sucks...
"and blew it. At least it’s good to know that we can take our robotic overlords at something, anyway."
Well now we know what to gamble controll of the world with! When the robot overlords begin to rise we shall challange them to a game of bowling for the fate of all the human race! (Except Martha Stewert. They can have her.)
Score one for H.U.A.R.
hmm i thought if you could save a strike you could get it repeated after that. Many things change but a near perfect strice should stay a good one even if many factores differ from frame to frame.
You'd think it'd be easier to make a better bowling robot. I mean, it's played indoors, so there's no wind. The alley is always the same, and the pins are always aranged the same way. All you'd need to do is set up the robot to do a strike, then merely repeat those peramiters until it gets a 300.
Actually, I've recently learned that lane conditions change pretty dramatically from day to day and probably even from frame to frame. All bowling lanes are oiled using a particular oil pattern. If you look at the PBA website, for example, you'll see that there are 5 or 6 specific oil patterns (each named, hilariously, after an animal) that must be used at all PBA events. They also keep separate stats on each bowler's performance on each oil pattern, since some patterns are better suited to specific play styles than others. The most consistently good bowlers are the ones who can adapt their style to these constantly changing conditions. My company recently made a bowling game (see URL), so I've been doing a bunch of research on this stuff.
Yes, the lanes change from frame to frame, especially with the new balls being used. The balls are so pourous, they soak the oil up as the go down the lane, so the lanes will "break down" the more they are bowled on. Depending on the house conditions, this could be slowly, or very fast. But as a general rule of thumb, in a three game series, you will have to move 3-8 boards, because the lanes will hook more later in the series.
Then there is humidity conditions, synthetic or wood lanes, type of lights (tungsten, florescent, etc...) amount of smoke (yep, effects the oil), etc.... Hell, a robot (or most programmers) wouldn't know this, but a good bowler will know that the closer you are to an outside wall of the building, the faster the lanes will break down.
Then we can discuss the machanics of the pin setters, but you would be bored. But we can leave it at the fact that they are not 100% accurate and can routinely place each pin up to 1/2" off spot. Multiple that by 10 and figure out how different each rack can be.
It is quite simplistic to say that since it is played indoors, that nothing changes. A bowling lane changes from hour to hour about as dramatically as a golf course does.
Throwing the ball is the easy part in bowling. Knowing where to throw it when is where a good bowler steps apart from the pack (especially on Sport conditions, not the slotted conditions most league bowlers bowl on, where you just throw the ball anywhere and it will hit the pocket.) A good bowler is learning from each shot and is adjusting the next one based on the knowledge of the first one.
If I were to design a machine to actually beat a human bowler in a 3 game match, it would not only have to be able to throw the ball, it would have to have cameras to detect angle, speed, rotation, trajectory and reaction. And it would have to be able to take all that information, store it and make adjustments based on the past balls it has thrown.
I am pretty sure this machine can't do that yet. :)