Cubestormer - Fastest Cubing Robot

[OregonTrail]

Embedding Disabled.





.

 

[Toad]

Tis very impressive... I guess it just calculates an optimal or nearly optimal solution with some nice programming...

 

[4Chan]

That 4 second solve was done with all pieces oriented. >.>
Someone scrambled it with less than 10 moves using half turns.

 

[Flicky]

That 4 second solve was done with all pieces oriented. >.>
Someone scrambled it with less than 10 moves using half turns.

Well done.

The video clearly says when it does a full solve. The short scrambles are just a demonstration of the robot.

 

[4Chan]

The video clearly says when it does a full solve. The short scrambles are just a demonstration of the robot.

I am aware of this, I obviously watched the video.

I was just making a comment.

 

[Dene]

That is SO COOL .

 

[ben1996123]

Wow cool!

 

[Muesli]

Woah.

I-i-i...

*sighs*

 

[CubesOfTheWorld]

they give the robot no inspection time

 

[Olivér Perge]

And I was building cars and houses from LEGO...

Very impressive job! 10-12 seconds for a Robot, are you kidding me?

 

[scottishcuber]

What happens if it encounters a lock up?

 

[OregonTrail]

What happens if it encounters a lock up?

Lockups only happen when one turns lazily enough to misalign the mechanism. I doubt that would happen to this robot.

If it did, the program might detect if the motor completed its rotation (based on the internal tachometer), and compensate if it did not.

 

[Edward]

:O
We can't let some machine surpass sheer will power and dedication...

Must Get Faster...

 

[ZamHalen]

Why are all these robots working and mine didn't lol. Now I feel like I suck(not at cubing but at building robots).All i tried to do was build the tilted twister and i failed at it(ok not really I just didn't have enough pieces in the kit we have at school).

More seriously though that robot's insane!!!!!

 

[brunson]

Can anyone reproduce the scramble and the solve? I'd like to know if the solve is optimal, because there's essentially no time for calculation. I guess it could be solving with a human method, the TPS is crazy fast, but I think it was less than 25ftm for the "random cube".

 

[Kian]

Can anyone reproduce the scramble and the solve? I'd like to know if the solve is optimal, because there's essentially no time for calculation. I guess it could be solving with a human method, the TPS is crazy fast, but I think it was less than 25ftm for the "random cube".

The TPS is not fast at all compared to a good human solver. It's not using any method I'm aware of.

Frankly, I would like to see some more proof that this is legitimate before I believe it at all.

 

[That70sShowDude]

 

[ianini]

That's so awesome! Does any one now a days solve using an optimal move method for speedsolving?

 

[qqwref]

:fp at everyone who doesn't realize it's running CubeExplorer and solving with that method (and that clearly this is not a method a human could use). If you're wondering where the processing power comes from, the computer part of it is probably running on that laptop over there.

Watching the full solve (the 10.75) it seems like it spends the first few seconds just doing an x4 y4 or something similar to scan all the faces, and then executes a 20-move solution. Looks like a completely legit CE robot to me. Very nice job, whoever made it

 

[JTW2007]

Why would it need to rotate mid-solve? Isn't rotation normally just to make algorithms faster to execute for human solvers?