Darts.

[Pete.]

cut down but interesting none the less

Must have been programmed by a football player going by the amount of showing-off it's done from getting one in the net

 

[daleyd]

I've never tried to build one tbh

BUT this is how I would do it

program the board layout into the computer system
overlay the board layout over the image being drawn into the computer vision system
Calibrate the arm (much as human would do), in computer terms you would aim it at every point on the board and do test throws to confirm calibration and use different brands of darts to account for different weights.
then its just a matter of making sure the robot is on its mark, that the vision system doesn't get impaired by smoke/beer etc.
After that you would just program the sequence and hit go

As @pdg said though, no one is going to pay the development costs (time to build, time to calibrate, time to test), its perfectly doable

Robotic arm wrestler...thats a different kettle of fish due to the issue of not breaking your opponents arm or crushing their hand.

(Phalanx also has a missile armed sibling btw)

Don't forget phalanx will hit a target with a small cross sectional area (smaller than a dartboard) which is travelling at Mach 2 while being fired from a moving platform. From "about" a couple of miles away.

IIRC they are generally on target within half a dozen rounds, then they move onto the next biggest bit of shrapnel.

 

[rtbcomp]

The flying shuttles of Victorian cotton mills came pretty to close to firing a missile to hit the same spot every time.

 

[daleyd]

The flying shuttles of Victorian cotton mills came pretty to close to firing a missile to hit the same spot every time.

Weren't they an 80's cappella band?

 

[WorkshopChris]

I think PDG hit the nail on the head with this statement,

The two videos posted are 'bloke in a shed' level of robotics.

Think what would happen if there was actually an application that would justify real funding.

The robots in the videos are very basic old tech.

They prove the concept so unless someone wins big on the lottery and goes crazy the same day i doubt anyone will spend time repeating the experiment, when they could demonstrate something more complex.

 

[daleyd]

Getting a dart in the same place every time is the (relatively) easy bit. The tricky part is adjusting where the next dart has to go because the bed is blocked, adjusting the angle you throw at to go in a different way, pressuring your opponent etc - robots are great at repeatability it's the "thinking" part that they can't really do - everything has to be programmed as an algorithm. (I know this is probably out of date and at some level machines are capable of a form of intelligence but I'm talking about general tech level not NASA type stuff)

So in answer to the first post, the mechanics of getting darts in the same place every time should be easily doable. Winning a darts match against a decent human player - probably not likely with current technology. (At least with a machine mimicking a human throwing action as there are just too many things to adjust,move etc, if it was a no holds barred "get the dart into the board any way you can" affair I reckon a machine might have a good chance)

 

[wookie]

Tell you what - show me proof that it's impossible.

Show me a robot failing to throw a dart accurately.

I'll believe it's impossible when I see it being impossible.

Ah, it's virtually impossible to prove an negative but a heck of a lot easier to prove a positive.

Example...

If you consider drug testing athletes proving a positive is fairly easy, you take a sample and run some test for substance "x" and get a positive result which presuming the test is legit then it means he/she cheated by taking substance "x".

However how do you prove a negative ? Well even if you test for all known substances and each test comes up negative that doesn't prove the athlete isn't cheating, only that either they aren't cheating or they are and have successfully masked the substance from your test or even they are taking some substance that you don't know about and therefore don't have a test for.

But I think I'm going a bit fftopic: here.