h a l f b a k e r yI think this would be a great thing to not do.
add, search, annotate, link, view, overview, recent, by name, random
news, help, about, links, report a problem
browse anonymously,
or get an account
and write.
register,
|
|
|
This is applicable to a variety of different situations, including cables, wool and knitted garments.
A needle attached to a loose end of a thread containing a small computer and a pair of bearings and cameras, one at each end. It's a given that the thread is at the back of the needle. The needle
swivels round on its bearings to detect the route followed by the thread to the nearest tangle, i.e. the point at which it passes into a loop. The angle of this point is determined by parallax between the two cameras. The bearings are swivelled to orient the needle to face at a precise right angle to that point, then the needle rolls backwards on them until it detects that it and the thread have passed entirely through the tangle, probably by sonar. Once this has happened, the process begins again for the next tangle.
So, i've got the basic ideas here with the central, hard problematic knot of how to detect where the last tangle is. I'm hoping that some of you will help me untie that last little bit. A brainless solution might involve the needle being steerable by remote control with a VR headset allowing the user to do it themselves.
[link]
|
| |
i was about to suggest that the needle spool on the thread as it goes backward on it, but remembering my last untangling sessions, i realized that it sometimes is impossible to pass a spool back through a loop. |
|
| |
I guess the needle would have to do what a human does in untangling a thread - freak out and buy a new one. Or, alternatively, go back through the loop, pull the string through the loop, and go through the next loop. |
|
| |
You could maybe pull it into a reel, but it'd be getting bigger all the time anyway, making it harder for it to get through the loop, particularly if it were knotted rather than just tangled. It could possibly be pulled into the body of the needle. I still think the biggest problem would be identifying the position of the next loop. |
|
| |
The robot would probably have to be two-part: one reel, and one climber - the climber simply goes back some way (determined by [thickness of thread] x [stiffness]x[some factor]), 'gives' the thread to the reel, goes back, pulls in the thread from the reel, until it has the end, then gives the end to the reel again, etc. that way the robot would not have to know about the loops, sometimes it would solve 3 loops in a row, sometimes none, but in the end all would be solved. |
|
| |