Idea for electronic cube

[Joël]

Hey guys,

I've been thinking about how hard it would be to build in sensors in all of the 6 centre pieces of a cube that can detect both clockwise and anti-clockwise movements, and connect them to a computer (possibly through some sort of wireless connection). It would be really cool if that was possible; you could get some very detailed information about the breakup times of your solves for example...

Am I the first one to think about this? I'd guess not...

Any people who are good with electronics and able to build something like this?

Bye!!

 

[miniGOINGS]

I think it would be easier to implement it into the edge pieces. You could have sensors that would show when 4 edges go past 4 other edges during a turn.

 

[Cyrus C.]

I think there was a thread about this, except so it would appear on a tv/computer screen. The general consensus was, with the current technology it would either weigh a lot, be too big, turn very bad, cost too much, or a mix of the four.

 

[Forte]

I thought about this a bit once (not making it, I suck at technology >_>).

My idea was like:
-blank cube with sensors and accelerometer
-screen with cube image, and does scrambles, etc.

Don't know how feasible it would be in practise though

 

[daniel0731ex]

or you could use a very strong elastic that ties each cubie to the core, and as you twist it the elastic binds itself in a certain way. the cube will then solve itself if you release it.

 

[miniGOINGS]

I think he means putting this into a cube so that you can download a detailed profile of your solve.

 

[kunz]

just put small wheels (like in a ball mouse) on the core up against the center piece and have a wireless (a wire would get in the way) link to a computer and have a 3d cube like gabbasoft (or anything really) but it probably would cost a lot

 

[Zarxrax]

Yea I made a thread about something like this before.

But recently me and a non-cuber friend were discussing it some more, and came up with a different design which might work better.

- Instead of stickers, you could have panels which use some kind of "e-ink" tech which can display various colors.
- Somehow fit a bluetooth transmiter into the cube somewhere, probably in the core
- have small batteries in the center pieces
- have metal connectors along the edges of all the pieces, so power can travel from the centers to the edges and corners.

Using the bluetooth transmitter, it could interface with a pc application to reset the cube to a clean state, or randomly set up cases like PLLs.

It's different from your idea, but both ideas could actually work together.

 

[Joël]

I think there was a thread about this, except so it would appear on a tv/computer screen. The general consensus was, with the current technology it would either weigh a lot, be too big, turn very bad, cost too much, or a mix of the four.

I refuse to believe it would weigh a lot or be too big, especially in a time where mobile phones are replacing our brains. I also don't see why it should turn very badly. I think if well designed, that shouldn't be a problem also.

The only objection I would agree with is that it would cost a lot, but it's certainly doable.. I know Rubik's made their own very expensive timer once, and that was very expensive too, but they did mass produce that too...

 

[miniGOINGS]

Either I'm completely nuts, or everyone is completely missing the point of this. It would function just like a normal 3x3, except a computer would be able to trace when moves are made.

 

[Joël]

Yea I made a thread about something like this before.

But recently me and a non-cuber friend were discussing it some more, and came up with a different design which might work better.

- Instead of stickers, you could have panels which use some kind of "e-ink" tech which can display various colors.
- Somehow fit a bluetooth transmiter into the cube somewhere, probably in the core
- have small batteries in the center pieces
- have metal connectors along the edges of all the pieces, so power can travel from the centers to the edges and corners.

Using the bluetooth transmitter, it could interface with a pc application to reset the cube to a clean state, or randomly set up cases like PLLs.

It's different from your idea, but both ideas could actually work together.

I personally would not like lights on my cube... I only want it to register moves.

 

[Zarxrax]

I personally would not like lights on my cube... I only want it to register moves.

But if you are going to start down that path, you may as well take it all the way, right?
I think what you are wanting might be possible with just a nice webcam and some intelligent programming. Maybe it could try to trace colors around the cube, or just look for your pauses and such, and make intelligent guesses about your timings.

 

[kunz]

now that i think of it there's already plenty of free space inside of the cubies and there is cheep tracing tech... wouldn't really be that hard

 

[Anthony]

I know Rubik's made their own very expensive timer once, and that was very expensive too, but they did mass produce that too...

Yeah, idiots like me bought it for $100 thinking it would be worth it. ><
Although, in my defense, I was using a stopwatch at the time and had no idea what a StackMat was.

 

[martin8768]

built in timer, wireless timer/tracking to a computer, rechargeable battery, OLED tiles that can be set to any color you want. please make this maybe?

 

[wing92]

I think what you are wanting might be possible with just a nice webcam and some intelligent programming. Maybe it could try to trace colors around the cube, or just look for your pauses and such, and make intelligent guesses about your timings.

This idea seems interesting to me. People can do weird things with webcams, it seems like following a colorful dot would be easier than other stuff they do.

Problems I see:
It would have to know the scramble or watch you scramble it (not too big of a deal)
Hands could get in the way of turns (could be a big deal but maybe fixable)
Alternate color schemes might be tricky. Also, different colored plastic might mess with the camera's recognition of colors, along with lighting.
The program would have to work pretty quick in the case of some fast algs.

And, the biggest problem: Somebody would have to put in the time to write the program. It's a good idea and definitely doable but I'm not sure if I could see anyone actually doing it.

 

[Cride5]

This would be great for FMC! Just do your solve and the system records your solution, automatically removing rotations and reversed moves.

@Joël your idea for the design is probably on the right track. A single axis accelerometer in each centre piece would probably sufficient for detecting the rotation. I think the main difficulty would be transmitting the information back to the computer. Bluetooth seems like the obvious choice. Either each centre piece needs to communicate electronically to a single transmitter, or 6 transmitters are used - each one broadcasting the state of each side.

If the system could be deployed as six chip+battery units (one for each centre piece), it would be pretty easy to install into your existing cube. Great idea! I would definitely buy something like this


EDIT: A quick search revealed that this kind of thing already exists - just search the term "Wireless Accelerometer". It looks like using a number of them simultaneously shouldn't be a problem. The smallest one I found is a 25mm square (5mm deep):
http://www.medibix.com/runsearch.js...tsort=manufacturer&view=sku&product_id=456675
http://www.microstrain.com/g-link.aspx
The biggest you would be able to fit into a centre piece would be about 15mm x 15mm, so these are just a little on the large side. I'm not sure if a smaller one exists, but if not we could talk to Guojia ... they seem to know a thing or two about cubes with big centres :rolleyes:

 

[qqwref]

This would be great for FMC! Just do your solve and the system records your solution, automatically removing rotations and reversed moves.

Would it be useful to have a simulator that did this? You could just have it apply the scramble (or inverse) and then do a solve with hi-games controls or something, and it would store your solution (including cancellations, and with no rotations or slice moves or whatever) or partial solution and tell you how many moves it was. A good FM-simulator would be able to store lots of solutions so you can choose the best or play with the ones you have. You could even have a function where you choose 3 stickers you want to do a commutator on and it would show you their position at every point in the solve, so you can try to find the best insertion.

Doing this type of thing with a webcam and physical cube would be quite difficult and a lot of work, but I think a FM-simulator would be very doable, especially if you already have code for the simulator itself.

 

[Cride5]

This would be great for FMC! Just do your solve and the system records your solution, automatically removing rotations and reversed moves.

Would it be useful to have a simulator that did this?

Oh definitely! At the moment I use http://alg.garron.us
While it allows you to turn the sides, it doesn't record the solution, so you have to type out the solution as you go. Sometimes I do too many moves without typing anything and forget how I got there - doh! It would be great if there was something like this, capable of receiving keyboard or mouse controls, which records your moves as you go. The other features you mentioned would be quite useful too, but then I'm not sure how far you can go with tools like this before you're breaking WCA regs.

 

[Tim Reynolds]

but then I'm not sure how far you can go with tools like this before you're breaking WCA regs.

E3) The competitor may use the following objects: paper and pen (both supplied by judge), 3 cubes (self supplied), unlimited coloured stickers (self supplied).

* E3a) Penalty for using other objects: disqualification of the solve.

In other words, if you're trying to not break WCA regs, you can't do anything like this...