Lego Robot Solves Any Rubik's Cube In 12 Seconds

kkleiner writes "Cube Stormer is the latest creation from Mike Dobson, aka Robotics Solutions, and not only is it made entirely out of Legos, it can solve any 3x3 Rubik's cube in less than twelve seconds. Often it can finish in less than five! This thing looks bad-ass and is incredible to watch."

Comments

Obligatory

[Anonymous Coward]

and not only is it made entirely out of Legos

It's "Lego". Lego bricks. Lego blocks. Lego. Not Legos. Lego.

Thanks

Re:Obligatory

[Waffle Iron]

Yes, non-pedants call each little piece of plastic a *LEGO*. How hard is that for you to understand?

Re:Obligatory

[H0p313ss]

Those that make fun of the first group and those who are utterly baffled at using a singular noun to describe something that is only interesting in the plural.

Like fish, and sheep...

Re:Obligatory

[mirix]

When you go hunting and fishing, do you bring back deers, elks, geeses, fishes? and drink some beers?

Cause I bring back deer, elk, geese and fish, and have a few beer. Well I would if I were a hunter.

Re:Not impressive

[SoVeryTired]

3X3 refers to the configuration of squares on each face of the cube. The 3x3 is the standard one that made us all so angry in the 80's.

Re:But can it...

[Anonymous Coward]

To make it unsolveable all you need to do is rotate one corner or one edge.

Re:Stickers

[JesseL]

And here's me expecting not to see stories from 9 years ago repeated.

http://slashdot.org/article.pl?sid=01/09/07/0133248

Wow. That site has more third party Javascript scr

[caluml]

Wow. That site has more third party Javascript scripts included than I've ever seen. It scrolls off my screen when listed in NoScript. *That's* why NoScript is good. :)

This Video is FAKE

[tonycheese]

Okay, it's not FAKE but it's completely and entirely dishonest. I can solve the rubik's cube in about 20 seconds over an average of 12 solves, so I have a thorough understanding of human speed-solving. Computers, on the other hand, would go for some idea solution that a human brain is not capable of producing. This is especially true since the robot in this video moves EXTREMELY slowly, about 1-2 turns per second on average. Human hands can EASILY sustain 3-5 moves per second. This computer, to solve the Rubik's cube in 2 seconds as in the first part of the video, or 4 seconds as in the second part of the video, would have to be able to solve the cube in 4-10 moves. The optimal solution for solving a rubik's cube has already been bounded at about 18 moves (look it up).

Still don't believe me? Start watching and replay the video from 30s onwards. Freeze the video when the timer starts at 0:00 and look at the cube, it is actually a single 90 degree rotation away from a fully solved state.

The 4s video beginning at 1:07 shows several rotations of the WHOLE CUBE without making any actual moves, then does 4 turns and solve it, which means that it wasn't anywhere near a scrambled state to begin with.

More evidence that it's fake? Is there any information on this other than a 2 minute video on youtube?

Re:This Video is FAKE

[newcastlejon]

The optimal solution for solving a rubik's cube has already been bounded at about 18 moves (look it up).

Only in the worst possible configuration of the cube. 18 moves can't be the lower bound for every cube, because there exist many configurations that can be solved in less than 18. (Like the one you mentioned at 30s) If you'd read the rest of the wiki article you probably just consulted you would have seen that there even configurations that need over 20 moves too.

As for turning the cube then solving it in 4 moves, look at the computer and note a single view of the cube. The machine has to determine the starting point before solving and perhaps one needs to see more than a single face to know enough to solve the cube.

I don't see how this is dishonest and you didn't say why you think it is. Is it because it uses a different algorithm than a human? Newsflash: Man invented the Machine to do things Man couldn't.

Re:This Video is FAKE

[tonycheese]

It's dishonest because for both the "2s" and "4s" solve of the cube, the cube was not fully scrambled. In fact, for the 2 second solve, the cube only had one single turn on it when the timer started. It is dishonest because he CLEARLY and obviously did not scramble the cube for both the 2 second and 4 second time. Look at the video at 30s and freeze it at the start of the timer and you'll see exactly what I mean. I can't honestly believe that you don't know what I mean by "dishonest" if you haven't done this simple task for me.

And yeah, you were right about the 18 moves thing, I was quickly looking for a number to back up my argument. The fact that 18 is actually lower than the optimal lower bound strengthens my argument instead of weakening it, though.

For the four moves, you need to see much more than a "single face" to solve a rubik's cube. When solving a cube, you do not look at faces, but rather look at the pieces that make up the cube: the "corners" which have 3 stickers on them, and the "edges" which have 2 stickers on them.

Re:This Video is FAKE

[ChinggisK]

He isn't claiming that it is fully scrambled in the 2s one - for the one immediately after he specifically says to get ready for a "full solve" and has big letters come up saying the NEXT CUBE is a "totally scrambled cube". The 2s was just a demonstration of the machine moving and is implied as such. You could maybe make a case about the 4 second one, but I'm not sure he's claiming that one to be a full scramble either.

Re:This Video is FAKE

[TheGeniusIsOut]

If you watch carefully, the "Full Solve", which states such and claims to be a "totally random cube (Honest)", takes 10.75 seconds, including inspection. The 2.01 second solve is a demonstration of the MINIMUM time required for "inspecting and making one twist" on an unsolved cube. It is the blogger, and not the video, who makes the claims of solving in 2.01 seconds, and while it technically is a solve, the inventor rightfully does not claim such. The world record human solve of 7.08 seconds is not including the untimed inspection period. I would not consider this a dishonest video, since the video does not claim anything but the 10.75s solve to be a real solve, which by the rules of the second video's competition, would actually be an 8.74s solve....

Re:This Video is FAKE

[newcastlejon]

It's dishonest because for both the "2s" and "4s" solve of the cube, the cube was not fully scrambled. In fact, for the 2 second solve, the cube only had one single turn on it when the timer started. It is dishonest because he CLEARLY and obviously did not scramble the cube for both the 2 second and 4 second time.

This isn't dishonest. I watched the video and saw that the machine can solve a trivial problem in one move. The video didn't disguise this in the slightest. You can see a more complex configuration being solved elsewhere in the video and this obviously takes more time.

Look at the video at 30s and freeze it at the start of the timer and you'll see exactly what I mean. I can't honestly believe that you don't know what I mean by "dishonest" if you haven't done this simple task for me.

At 0:30 I see a cut from one sequence to another. I didn't think I was watching a real demonstration until I saw a start-to-finish run without any cuts.

And yeah, you were right about the 18 moves thing, I was quickly looking for a number to back up my argument. The fact that 18 is actually lower than the optimal lower bound strengthens my argument instead of weakening it, though.

No, it doesn't. Your original argument seemed to be that 18 moves is the lower bound for solving any particular configuration of a Rubik's Cube; this is just wrong.

I can show you at least a dozen possible configurations that need only one move to solve. If you include an already solved cube then the lower bound becomes zero. The upper bound is the significant one because for any starting position you can solve the cube in x moves or fewer. However, the upper bound means there is/are configurations that cannot be solved in less than x moves. If this was your point, then so what? No cube can be solved in fewer moves than it took to set it up, and the fewer moves that took, the lower the upper bound for solving it.

For the four moves, you need to see much more than a "single face" to solve a rubik's cube.

I didn't suggest that the machine looked at only one face, I said it could only look at one at a time. The 3-4 moves are so the machine can look at more than one face to fully determine the state the cube was in.

However, When solving a cube, you do not look at faces, but rather look at the pieces that make up the cube: the "corners" which have 3 stickers on them, and the "edges" which have 2 stickers on them.

What particular parts of each face are important is - from a practical standpoint - irrelevant. The machine still needs to look at more than one face irrespective of what bits the algorithm works with.

Re:How they are doing it?

[Paiev]

Just as a quick note, the page you linked is unofficial world records and hasn't been updated in a while. The official fastest times are here

Before someone claims dishonesty, all these solves were performed in competition with judges observing,

Re:How they are doing it?

[ElectricTurtle]

Yeah, I don't know anything, I just watched the private lives of national competitors. I judge them lacking. That is my prerogative.

Also, Scrabble lacks depth at higher levels. National competitors learn spellings but not necessarily meanings, which completely bastardizes the entire point of language. National Scrabble competitors essentially live or die by the number of combinations of seven letters they know are valid. To me that is a disgusting travesty, and quite frankly after watching Word Wars I discovered I had a lower opinion of tournament Scrabble players than I did before I knew anything about them. Once again, my prerogative. I don't know you, and if you're different, great, but you're not the only Scrabble player, and I'm more keen to trust what I've seen than you, quite frankly.

Re:How they are doing it?

[Paiev]

It depends on your perspective. You're right in that competitive Scrabble isn't about knowing meanings of words, and this does bother some people. You can view Scrabble as not being about words so much as about combinations of letters. The game is then about a mix of anagramming and strategy (it was this strategy that I was talking about when I referred to depth; while living room Scrabble games usually feature little strategy, competitive players take into account many factors when choosing a play).

Scrabble isn't a game about words, it's just a game that uses them. It's about anagramming and about strategic decisions. It's no coincidence that Scrabble players tend to be on the more math-y side. If you shift your view of the point of the game, you might lose some of your disgust.