World Record? Robot "Sub1" :)

Berd said:
Amazing! Wca scrambling bot anyone?
Click to expand...

Not this one I suspect, considering how long it probably takes to insert and remove the cube and that the center piece has to be modified for the robot to hold it! ;)

EDIT:

Oh and BTW, I added a scramble option to the latest version of the MindCub3r software which generates a random position (rather than random move) based scramble so is probably good enough for practice if not competitions. Of course mechanically it is probably far too slow for most of you guys during pactice! ;)

If you are interested see: http://mindcuber.com/mindcub3r/mindcub3r.html#Operation
 
Last edited:
Oh, this was a different 1-second-ish robot? :p Actually, when I saw this, I thought it was the Flatland/Rose one with a new chassis. Very cool that there are multiple super-fast robots.

I think when the times are this close it is really unacceptable to use a single solve, though. The position of the cube matters a lot - it will affect both solving time and the amount of time it takes to do the move sequence. I'd really like to see at least an average of 12, or even an average of 100. Even with a human scrambling the cube, I would not expect 100 solves to take more than an hour.
 
qqwref said:
Oh, this was a different 1-second-ish robot? :p Actually, when I saw this, I thought it was the Flatland/Rose one with a new chassis. Very cool that there are multiple super-fast robots.

I think when the times are this close it is really unacceptable to use a single solve, though. The position of the cube matters a lot - it will affect both solving time and the amount of time it takes to do the move sequence. I'd really like to see at least an average of 12, or even an average of 100. Even with a human scrambling the cube, I would not expect 100 solves to take more than an hour.
Click to expand...

I agree!

And also it needs to be defined and confirmed exactly what is timed...

e.g. is it ok for the camera streaming images to the computer before the timer starts? How do they ensure the timer starts before the camera starts sampling the image that is actually used for image analysis rather than using one that sampled just before the timer started and was buffered in the system? On CUBESTORMER 3 for example, we found by experiment that we believe there was approximately a 2.5 frame time delay through the camera system from the image actually being taken by the camera and it becoming available to the software and we included this in our timing method. When does it finish?

Is it when the last pulse has been applied to the stepper motor controller (what if you are not using stepper motors like CUBESTORMER 3) or should it have some way of seeing that the cube has actually stopped turning? Or should it be only to the point at which the last face is within 45 degrees of solved (since that defines "solved" in human competitions)...

Maybe I'm being pedantic, but certainly a camera frame time could be longer than the difference between 0.900s and 0.887s so even with an average time to even out cube scrambling differences the times could be close enough to be affected simply by how the timing system works! :)
 
Well think of the battle between humans and computers in chess. I guess that means someday a human cuber will beat the robot..? Like chess they'll have human vs. robot championships.
 
You must log in or register to reply here.

Similar threads

E
Robot solves Megaminx in 7.647 seconds (WORLD RECORD)
2
Replies
22
Views
1,535
iguana
iguana
IAssemble
Cubestormer 3
2
Replies
25
Views
7,599
Kirjava
Kirjava
IAssemble
  • Locked
[Unofficial] CubeStormer II sets Guinness World Record for fastest robot solving a Rubik's Cube
Replies
15
Views
3,461
abunickabhi
abunickabhi
IAssemble
  • Locked
[Unofficial] CubeStormer II - robot solves Rubik's Cube faster than human world record
2 3 4
Replies
70
Views
17,472
RoboticSolution
RoboticSolution
AK51
2.51 world record used 33 moves, it shall be 28 moves using AK-cubie method
Replies
0
Views
268
AK51
AK51
Back
Top