Would it be possible to build a "self-solving" rubik

https://youtu.be/ixTddQQ2Hs4

>>1032396
but, that's the robot who is moving the cube.
I'm talking about a cube that doesn move alone.

>>1032405
does*

It could be possible if we had nanotechnology to put in the center 4 motors and something like an arduino board. But i think its an unnecessary thing, cause the cube is for help memorizing and understand a little more of movement.

File: damnit anon.jpg (100KB, 796x800px) Image search: [Google]

100KB, 796x800px

>>1032408
oh ok, i just imagined it because it would look pretty cool.
If someday i have access to nanobots i'll try to make it real.

File: cdc.jpg (45KB, 546x366px)

45KB, 546x366px

>>1032409
Have you been watching this guy:
https://www.youtube.com/watch?v=rzkAN3oQjXc

>>1032392
Rather than a moving cube how about six screens in a cube configuration with animated "sides", it would be one hell of a job for the person coding it but I could see it working.

>>1032503
It's called a touch cube.

>>1032504
And apparently it can solve itself, like op wanted.

File: received_1686848301578183.jpg (7KB, 199x197px) Image search: [Google]

7KB, 199x197px

>>1032532
Yeah but it doesn't move like it's alive or magic.

>>1032392
disassemble the cube, study the mechanics, then study it some more, then some more and then when you completely understand how it works come up with a way to drive motors moving six sides left/right on all axises

the pieces lock together and slide around a central ball, what about putting windings in each piece and then seting it up so you can energize the windings making each individual piece magnetized, reverse polarity's on specific pieces to start them rotating away from other planes... or perhaps a rare earth magnet in the center of each piece and somehow come up with a way of layering windings on the central ball that would allow you to rotate each of the individual working planes of the cube?

it'd be expensive to build, and I suspect you'd need a lot of charge to run one properly, but hell, might be worth a shot and I think it's more feasible than using tiny motors of some kind....

oddly it may help to specially make an oversized cube... since the center can be effectively hollow you aren't really increasing the weight of the pieces and the more room for the windings you have in the center ball I suspect the better off you will be

>>1033271
hmm..

on this line of thought you'd need 4 sets of winding paths... 3 of them would be parallel with each other and would allow you to manipulate the planes of the cube, the 4th would let you spin the ball relative to the cube and would need to be perpendicular to the other 3... (you need to be able to move all 3 axis of the cube, the 3 parallel let you manipulate in say the x axis, the 4th lets you rotate along the z axis, and to get to the y axis you rotate all 3 x axis rings at once turning the ball 90 degrees relative to the outside then turn it another 90 on the z axis)

maybe look up the windings used in brushless wound stator dc motors.. you can adjust speed on those, run them really slow, and reverse direction... real problem is getting all the battery to do this without needing some kind of plug in cube...

File: 00657-01.jpg (23KB, 350x350px)

23KB, 350x350px

>>1032392
I can imagine how to make it move, but I can't imagine how the cube would figure out what state it is in.

Just make the cube bigger and put geared motors in each of the joints.

Sensing state will be interesting, you would probably have to do it optically by putting some LEDs and photodiodes to read little 3 bit barcodes. Heck maybe you could do the trick where you use an LED as a photo proximity sensor to read the barcodes.

One could probably make a regular sized rubiks cube motorized if one was willing to buy high precision motors with gearboxes made by goddamn swiss watch makers.

Keep an eye on developments in ultrasonic motors, they have the potential to be very torquey and tiny

>>1033371
You start with the cube solved and have it remember what turns are made to it so you don't need color sensors or anything like that.. Then either reverse the moves or run it through a legitimate solving algorithm

File: tumblr_nnyfsbrFjX1rsi8pxo1_500.png (98KB, 500x505px) Image search: [Google]

98KB, 500x505px

You could tire rubber bands to the back of each piece and the other end of the band to middle piece. When you mix up the cube, the rubber bands stretch and when you let the cube go, it will twist back into shape.

>>1033553
wat

>>1033553
the scary thing is that if you had an infinitely stretchy rubber band, and no turning friction for this to overcome then it would work as long as the sides didn't jam up on each other... as you twisted the cube more the most recent rubber band would be on top...