Online since 2002. Over 3300 puzzles, 2600 worldwide members, and 270,000 messages.

TwistyPuzzles.com Forum

It is currently Wed Jul 23, 2014 3:34 pm

All times are UTC - 5 hours



Post new topic Reply to topic  [ 20 posts ] 
Author Message
 Post subject: Robotic Self Solving 3x3
PostPosted: Tue Apr 08, 2014 7:19 pm 
Offline
User avatar

Joined: Mon Nov 16, 2009 10:50 am
Hi TP,

I don't know if anyone even still uses this site but I figured at least someone would see this. So I built this cube that solves itself. I tried making it as simple as possible but its still sorta complex. So basically I put motors under 5 of the centers and ran the wiring through one face. The bottom face just doesn't turn. I couldn't think of a reasonable way to do that. I was an idiot when I first started and thought it would be smart to use stepper motors. The reasoning was the consistency in its turning but steppers are ridiculously weak. The whole design honestly is flawed just because it's so hard to control the individual pieces. I couldn't think of a smart way for the cube to register turns so I made it so you turn it with switch inputs. I also decided to make the program so that it logically finds a solution rather than reversing the scramble because it could very well have access to that. So The whole logic is done on my computer which is toggled to the arduino(seen below). The arduino just takes computer commands and uses them to run the five stepper drivers. The results were not pleasing. It cant get past one turn. It would be really hard to make one of these that actually works, but it's probably possible. Oh and because of the ridiculous size of the stepper motors I had to scale the design way up. It's about the size of a cantaloupe (125mm). Definitely no V2 coming from me that was expensive :shock:


Attachment:
20140408_194652.jpg
20140408_194652.jpg [ 1.8 MiB | Viewed 1937 times ]


Attachment:
20140408_194606.jpg
20140408_194606.jpg [ 2.11 MiB | Viewed 1937 times ]


Attachment:
20140408_194554.jpg
20140408_194554.jpg [ 1.78 MiB | Viewed 1937 times ]

_________________
"If I have been able to see further, it was only because I stood on the shoulders of giants" Isaac Newton


Top
 Profile  
 
 Post subject: Re: Robotic Self Solving 3x3
PostPosted: Wed Apr 09, 2014 1:11 am 
Offline
User avatar

Joined: Thu Dec 21, 2006 5:32 pm
Location: Bay Area, CA
Very cool!

Dave

_________________
Image
LitwinPuzzles.com has info on my puzzles.


Top
 Profile  
 
 Post subject: Re: Robotic Self Solving 3x3
PostPosted: Wed Apr 09, 2014 2:45 am 
Offline
User avatar

Joined: Tue Sep 11, 2012 1:51 pm
Location: Bedford, England
yummyyummypbj wrote:
I don't know if anyone even still uses this site


:shock:

Very cool though, I'd love to see a video of it in action.

_________________
My youtube channel: http://www.youtube.com/user/martywolfman?feature=plcp Now in 1080p HD!

Latest videos: Unboxing Burgo's incredible Crazy 2 Face B4 kit, Calvins 2x4x6


Top
 Profile  
 
 Post subject: Re: Robotic Self Solving 3x3
PostPosted: Wed Apr 09, 2014 12:20 pm 
Offline
User avatar

Joined: Fri Feb 18, 2000 8:50 am
Location: chicago, IL area U.S.A
This is a great start! I was actually thinking the other day that I'm surprised no one has made a self solving Rubik's cube. I would imagine piggybacking off of this, eventually a self contained self solving rubik's cube could be made.

Just out of curiosity, is it possible to solve from any position using only 4 axes? (I think this may have already been discussed, but I don't recall for sure.)

-d


Top
 Profile  
 
 Post subject: Re: Robotic Self Solving 3x3
PostPosted: Wed Apr 09, 2014 2:05 pm 
Offline
User avatar

Joined: Thu Dec 02, 2004 12:09 pm
Location: Missouri
darryl wrote:
Just out of curiosity, is it possible to solve from any position using only 4 axes? (I think this may have already been discussed, but I don't recall for sure.)
Hmmm... I think I know what you are after but the question isn't worded very well. First the Rubik's Cube has 3 axes or 6 faces. Here yummyyummypbj only allows 5 faces to turn. And yes this still allows all states of the Rubik's Cube to be reached. This isn't true for a Super 3x3x3 which takes into account face center rotation but that is a minor point. I would love to see a self solving Rubik Cube someday. Better yet would be one that got mass producessed and was available for purchase.

And in case you were wondering if only allowing 4 faces to turn was still a fully generalized Rubik's Cube the answer is no.

Carl

_________________
-
Image

Image


Top
 Profile  
 
 Post subject: Re: Robotic Self Solving 3x3
PostPosted: Wed Apr 09, 2014 2:15 pm 
Offline
User avatar

Joined: Thu Dec 02, 2004 12:09 pm
Location: Missouri
One tip that might help with some of your issues. Since the stepper motors are connected directly to the face centers you want the face centers to me as large as possible to maximize the face centers ability to force the edges and corners to turn with them. There are two ways I see to do this:

(1) Go with a non-proportional cube which has very wide slice layers.

(2) If you want to keep the cube proportional (and who wouldn't) use the V-Cube mech shown here. It adds a lip onto the face center that extends into the edges and corners.

Image

Hope this helps,
Carl

_________________
-
Image

Image


Top
 Profile  
 
 Post subject: Re: Robotic Self Solving 3x3
PostPosted: Wed Apr 09, 2014 2:25 pm 
Offline
User avatar

Joined: Sun Jul 14, 2013 9:43 am
Wow. This is awesome. I imagine that it would be difficult to program the cube to remember how it has been scrambled, but not just solve it as a reverse scramble.

_________________
ȨZRĄ is almost my name


Top
 Profile  
 
 Post subject: Re: Robotic Self Solving 3x3
PostPosted: Wed Apr 09, 2014 4:11 pm 
Offline
User avatar

Joined: Tue Nov 26, 2013 4:57 pm
Location: Green Bay, WI
I would suggest adding a clicking mechanism for better alignment, instead of copying V-Cube.

_________________
I hope you have a cloudy day today
I hope you have a miserable day
I hope everyone is a pain
I hope it rains on your parade
I hope you have a lousy day today

Just Kidding! Have a great day!

Sparta? THIS IS MADNESS!!!!!


Top
 Profile  
 
 Post subject: Re: Robotic Self Solving 3x3
PostPosted: Wed Apr 09, 2014 4:34 pm 
Offline
User avatar

Joined: Fri Feb 18, 2000 8:50 am
Location: chicago, IL area U.S.A
MODERATOR: I've moved the topic here since I believe it's more of a work in progress.

-d


Top
 Profile  
 
 Post subject: Re: Robotic Self Solving 3x3
PostPosted: Wed Apr 09, 2014 4:42 pm 
Offline
User avatar

Joined: Fri Feb 18, 2000 8:50 am
Location: chicago, IL area U.S.A
wwwmwww wrote:
And in case you were wondering if only allowing 4 faces to turn was still a fully generalized Rubik's Cube the answer is no.

Carl


Yes, I worded that poorly. I meant to say 'faces'. So one can not use four faces and get to every possible position is what it sounds like then.
Next question of course would be: How many positions are possible using turns on only 4 of the faces then? :)
I don't mean to get too off topic, but it does seem relevant since it could potentially simplify a design.

-d


Top
 Profile  
 
 Post subject: Re: Robotic Self Solving 3x3
PostPosted: Wed Apr 09, 2014 5:29 pm 
Offline
User avatar

Joined: Thu Dec 02, 2004 12:09 pm
Location: Missouri
E-Cubes Designer X wrote:
I would suggest adding a clicking mechanism for better alignment, instead of copying V-Cube.
No reason you can't add a clicking mechanism to the V-Cube mech.
darryl wrote:
MODERATOR: I've moved the topic here since I believe it's more of a work in progress.
I hope you are correct. The comment about there NOT being a version 2 from him makes me think he's given up on it as it stands and he's passing on the idea to others that may want to take a shot.
darryl wrote:
So one can not use four faces and get to every possible position is what it sounds like then.
That is correct.
darryl wrote:
Next question of course would be: How many positions are possible using turns on only 4 of the faces then? :)
I think that depends... are the two faces that aren't used adjacent or opposite? If adjacent it should be obvious that the edge between those two face centers could never be scrambled. If the faces are opposite its a little tricker to see but one limitation that would impose is that you could never flip an edge and return it to its starting position at the same time. There may be other limitations as well but that is the easiest one for me to see at the moment just moving pieces around in my head. To calculate the exact number of permutations for each would require some work in GAP that I haven't done yet. Maybe later...

Carl

_________________
-
Image

Image


Top
 Profile  
 
 Post subject: Re: Robotic Self Solving 3x3
PostPosted: Thu Apr 10, 2014 12:06 pm 
Offline
User avatar

Joined: Fri Feb 18, 2000 8:50 am
Location: chicago, IL area U.S.A
wwwmwww wrote:
darryl wrote:
MODERATOR: I've moved the topic here since I believe it's more of a work in progress.
I hope you are correct. The comment about there NOT being a version 2 from him makes me think he's given up on it as it stands and he's passing on the idea to others that may want to take a shot.


I think that depends... are the two faces that aren't used adjacent or opposite? If adjacent it should be obvious that the edge between those two face centers could never be scrambled. If the faces are opposite its a little tricker to see but one limitation that would impose is that you could never flip an edge and return it to its starting position at the same time. There may be other limitations as well but that is the easiest one for me to see at the moment just moving pieces around in my head. To calculate the exact number of permutations for each would require some work in GAP that I haven't done yet. Maybe later...

Carl


MODERATOR: Yes, it is kind of borderline which forum the post belongs in. I opted to move it here hoping someone else can continue the work.

Yes, I was referring to opposite sides not moving. I will have to think and work through what you said about the edge not flipping and returning to it's starting position. Thanks for the detailed reply!

-d


Top
 Profile  
 
 Post subject: Re: Robotic Self Solving 3x3
PostPosted: Thu Apr 10, 2014 3:45 pm 
Offline
User avatar

Joined: Tue Aug 11, 2009 2:44 pm
Cool idea. A self-solving cube would be a nice object to own, and fun (and challenging) to build.

Maybe you want alignment sensors of some sort between corners and adjacent edges? And power/sensor distribution with electrical rails so that the corners and edges can participate.

Also yes, ideally regular size, and ideally not connected to a computer. Probably it should have a platform with "scramble" and "solve" buttons; it scrambles as long as you hold down the scramble button.


Top
 Profile  
 
 Post subject: Re: Robotic Self Solving 3x3
PostPosted: Thu Apr 10, 2014 5:01 pm 
Offline
User avatar

Joined: Mon Oct 18, 2010 9:44 am
bhearn wrote:
Cool idea. A self-solving cube would be a nice object to own, and fun (and challenging) to build.

Now hook up one of those brain-control interfaces to this thing and that would be really cool and exciting. Solve it with your mind.

http://www.popsci.com/technology/article/2011-03/medicine-mind-using-eeg-patients-can-play-music-merely-thinking


Top
 Profile  
 
 Post subject: Re: Robotic Self Solving 3x3
PostPosted: Thu Apr 10, 2014 8:58 pm 
Offline
User avatar

Joined: Thu Dec 02, 2004 12:09 pm
Location: Missouri
bhearn wrote:
Also yes, ideally regular size, and ideally not connected to a computer. Probably it should have a platform with "scramble" and "solve" buttons; it scrambles as long as you hold down the scramble button.
I'd like the platform to have another button. Let's call it "clock". Have it apply the Hamiltonian Circuit for the 3x3x3 discussed here. Note that Hamiltonian Circuit only uses 5 faces so its perfect for this puzzle. Have the clock start at solved say on January 1st, 1970 at 00:00:00. (I believe this is know as Unix Time) And each move into Hamiltonian Circuit counts as 1 second. Then you could just let this run as a clock though current time. You'd never have to worry about the puzzle every repeating states. The issue of actually telling what the time is from just examining the cube though is left as an exercise for the owner. It would probably set the record as the world's most geeky clock. I'd rather have this then a clock like this anyday...
Image

Carl

_________________
-
Image

Image


Top
 Profile  
 
 Post subject: Re: Robotic Self Solving 3x3
PostPosted: Thu Apr 10, 2014 9:11 pm 
Offline
User avatar

Joined: Fri Dec 28, 2012 1:50 pm
Location: Near Las Vegas, NV
Slightly off-topic, but this reminded me of this which was posted quite a while ago. Is there any more progress on this? Eitan?

_________________
My Youtube channel
My Shapeways Shop


Top
 Profile  
 
 Post subject: Re: Robotic Self Solving 3x3
PostPosted: Fri Apr 11, 2014 10:01 am 
Offline
User avatar

Joined: Wed Jan 07, 2009 6:46 pm
Location: San Francisco, CA
I'd say, "Shhhhhhhh...." but I suppose the cat's out of the bag now... :roll:

That teaser was from Attempt #2 last summer. Attempt #3 is going very well. Parts arrive next Friday, so hopefully I'll post some progress by the end of April.

For now, let's keep this conversation on the topic of yummyyummypbj's attempt.

-Eitan

_________________
Eitan = "EIGHT-ahn"
Buy a Radio Cube 3! Only $150 at Eitan's Shapeways Shop
Check out my video: Twisty Puzzles a la Vi



Top
 Profile  
 
 Post subject: Re: Robotic Self Solving 3x3
PostPosted: Sat Apr 12, 2014 11:49 am 
Offline
User avatar

Joined: Mon Nov 30, 2009 1:03 pm
This is my test of how much torque is needed for turning different types of Rubik's Cube. It is just under 0.1 Nm, which still means that you would need either a big servo/stepper motor, or a small servo motor with a gearbox with a significant reduction. The bad part of the latter solution is that manual turning of the cube would damage the gear box.

It has been suggested to use half of a Dumbbell Cube to build a Robo Cube. The gear box and concentric mechanism would enable to power all six axles, not just five of them. Unfortunately, the friction of that gear box is, well ..., significant.

Perhaps we should start with a Robo Skewb first, e.g. based on Dumbbell Skewb. That one has only four axles.

Oskar

_________________
Oskar's home page, YouTube, Shapeways Shop, Puzzlemaster, and fan club
Image.


Last edited by Oskar on Sat Apr 12, 2014 12:25 pm, edited 1 time in total.

Top
 Profile  
 
 Post subject: Re: Robotic Self Solving 3x3
PostPosted: Sat Apr 12, 2014 12:16 pm 
Offline
User avatar

Joined: Thu Dec 02, 2004 12:09 pm
Location: Missouri
Oskar wrote:
The gear box and concentric mechanism would enable to power all six axles, not just four of them.
Yummyyummypbj's attempt powers 5 faces... not 4. And as long as one is dealing with a normal Rubik's Cube (not a Super Cube) powering 5 faces is sufficient to enable it to solve itself. Granted it would be nice if all 6 faces could be turned if a human could turn the faces as well but still only 5 motorized faces are needed if face center orientation can't be seen. That would be a nice way to force the puzzle to find a different solution and not just perform a reverse scramble.

Carl

_________________
-
Image

Image


Top
 Profile  
 
 Post subject: Re: Robotic Self Solving 3x3
PostPosted: Sat Apr 12, 2014 3:19 pm 
Offline

Joined: Mon Aug 18, 2008 10:16 pm
Location: Somewhere Else
Oskar wrote:
Perhaps we should start with a Robo Skewb first, e.g. based on Dumbbell Skewb. That one has only four axles.


A Robo 2x2x2 would only use 3. :)


Top
 Profile  
 
Display posts from previous:  Sort by  
Post new topic Reply to topic  [ 20 posts ] 

All times are UTC - 5 hours


Who is online

Users browsing this forum: No registered users and 6 guests


You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot post attachments in this forum

Search for:
Jump to:  

Forum powered by phpBB © 2000, 2002, 2005, 2007 phpBB Group