Robot beats Feliks record

Gus · **Posted:** Wed Oct 19, 2011 2:14 am

Our robot overloards will soon be here! A robot made from Lego Mindstorms has beaten the 3x3x3 record set by Feliks - see here for details. We are all doomed! The tin foil hats are useless! Run for the hills!

maarten · **Joined:** Mon Feb 02, 2009 3:50 pm

Does that count as a two-hand solve?

Burgo · **Posted:** Wed Oct 19, 2011 2:46 am

I didn't see it put it's hands on the stackmat afterwards, isn't that a 2sec penalty?
Cheers,
Burgo.

Gus · **Posted:** Wed Oct 19, 2011 3:11 am

maarten wrote:

Does that count as a two-hand solve?

Are humans with three or more hands barred from taking part in Speedsolving competitions?

Burgo wrote:

I didn't see it put it's hands on the stackmat afterwards, isn't that a 2sec penalty?

The stackmat is built into the robot. Perhaps Feliks could have one surgically grafted onto his body to improve his times?

alaskajoe · **Joined:** Mon Feb 06, 2006 12:52 am

This robot is good. Holy smokes!
Amazing build. Amazing achievement by the builders (humans btw) :wink:

RubixFreakGreg · **Posted:** Thu Oct 20, 2011 12:48 pm

Am I the only one who noticed those 5 seconds INCLUDE observation time ? :shock:

Luke · **Posted:** Thu Oct 20, 2011 12:51 pm

Very impressive. I'll be happy to bet in ten years time this would be considered a bad solve for a robot.

IlTallDude · **Joined:** Wed Jul 13, 2011 12:51 pm

Hmmm I wonder what the time is without inspection... but I believe the record for machines has to include inspection.

But it seems like it could be done faster with faster motors.... and maybe a guhong

KelvinS · **Joined:** Mon Mar 30, 2009 5:13 pm

IlTallDude wrote:

Hmmm I wonder what the time is without inspection... but I believe the record for machines has to include inspection.

But it seems like it could be done faster with faster motors.... and maybe a guhong

I was just wondering how the time breaks up between the 3 stages [inspection -> solution computation -> physical movement], whether there is any overlap between these stages [does it start rotating the layers even before the algorithm has been calculated?], and to what extent each of these 3 stages can be done faster, e.g., with faster processor, faster motors, and a faster cube, etc.

I wonder if there is any theoretical limit, based on friction, strength of materials, and the speed of light?

Jorbs3210 · **Posted:** Mon Oct 31, 2011 3:58 pm

KelvinS wrote:

I was just wondering how the time breaks up between the 3 stages [inspection -> solution computation -> physical movement], whether there is any overlap between these stages [does it start rotating the layers even before the algorithm has been calculated?], and to what extent each of these 3 stages can be done faster, e.g., with faster processor, faster motors, and a faster cube, etc.

Solution computation and physical movement are overlapped.

IAssemble · **Joined:** Wed Apr 06, 2011 7:59 pm

Jorbs3210 wrote:

KelvinS wrote:

I was just wondering how the time breaks up between the 3 stages [inspection -> solution computation -> physical movement], whether there is any overlap between these stages [does it start rotating the layers even before the algorithm has been calculated?], and to what extent each of these 3 stages can be done faster, e.g., with faster processor, faster motors, and a faster cube, etc.

Solution computation and physical movement are overlapped.

The software calculates the complete solution before the physical solve starts in order to find an almost optimal solution.

The time is divided roughly as follows:

1s inspection (involving re-orientation of the whole cube)
0.2s solution computation (slight overlap with recovery after cube reorientation)
~4s physical solve (starts only after complete solution is calculated)

Robot beats Feliks record

Who is online