A completely different technological approach for smart cubes

[FinnTheCuber]

No i dont think so it would just seem that closer to microwave frequency would be for rfid.

Edit: just did some research and it looks like 13.56 MHz is used. I learned something new!

 

[Wish Lin]

Important new update (2020/5/23)
Just finished my mid-term exam, and I have some good news and bad news for you guys:

Let's start off with good news:

The. Thing. Actually. Worked.
Yes, I build the first prototype and tested it. Although the whole setup was kind of crappy, it WORKED.

Bad news:
It turned out to be an impractical idea for speedcubing.

Detailed process:

I finished 3D-printing all the parts around April, then got all the electronics ready recently. It worked just as expected.
But I instantly found out that this idea is not good, here is why:

• The biggest problem is HOW ACCURATE YOU WILL HAVE TO TURN THE CUBE. Because my setup(I have tried every sensor/magnet setup I could think of) is already at the limit of the detection range of the Hall sensor, you have to turn the cube PERFECTLY ALIGNED WITH THE SENSOR plus no corner cutting at all. By now, if you are a speedcuber, you know it's a bust. If I try upgrading the sensor, the price just SKYROCKETS and you can get 2~3 356is with that price.
• The cheap Hall sensors' data fluctuate easily, adding to the fact that it is already at detection limit, makes the whole thing very unstable and inconsistent.

With this failure, I will say that all smart cube ideas regarding to an external frame is probably a bust(in my limited knowledge, of course). But I can't deny it's a success anyway, and I will NOT stop developing next generation smartcubes until it becomes a reality.

Frame of the(already disassembled) prototype:

A hall sensor stuck on it and its legs broke off when I tried to remove it, so I just leave it there. I recovered all the electronics BTW.

 

[brododragon]

Important new update (2020/5/23)
Just finished my mid-term exam, and I have some good news and bad news for you guys:

Let's start off with good news:

The. Thing. Actually. Worked.
Yes, I build the first prototype and tested it. Although the whole setup was kind of crappy, it WORKED.

Bad news:
It turned out to be an impractical idea for speedcubing.

Detailed process:

I finished 3D-printing all the parts around April, then got all the electronics ready recently. It worked just as expected.
But I instantly found out this idea is not good, and here is why:

• The biggest problem is HOW ACCURATE YOU WILL HAVE TO TURN THE CUBE. Because my setup(I have tried every sensor/magnet setup I could think of) is already at the limit of the detection range of the Hall sensor, you have to turn the cube PERFECTLY ALIGNED WITH THE SENSOR plus no corner cutting at all. By now, if you are a speedcuber, you know it's a bust.
• The cheap Hall sensors' data fluctuate easily, adding to the fact that it is already at detection limit, makes the whole thing very unstable and inconsistent.

With this failure, I will say that all smart cube ideas regarding to an external frame is probably a bust(in my limited knowledge, of course). But I can't deny it's a success anyway, and I will NOT stop developing next generation smartcubes until it becomes a reality.

Frame of the(already disassembled) prototype:

A hall sensor stuck on it and broke off when I tried to remove it, so I just leave it there. I recovered all the electronics BTW.

Could you use something to detect when a turn gets slowed by a magnet? Probably dumb but, who knows?

 

[Wish Lin]

Could you use something to detect when a turn gets slowed by a magnet? Probably dumb but, who knows?

Yes, but I can't tell the turning direction.

 

[Wish Lin]

A completely new approach is on its way!

 

[Wish Lin]

The details will be revealed later, but this time, being cheap is no longer the primary goal, this idea is about a possibly all-cube-smart-cube system!(Of course, I still want to make it as cheap as possible, but $50 or less probably isn’t possible)

(Sorry @Etotheipi ,I failed to make the cheap smart cube you have hoped for, but someone out there will definitely make it in the near future, probably using a different insight compared to mine!)

Currently discussing the idea with @Lelouch, who is experienced in neural networks.

 

[brododragon]

The details will be revealed later, but this time, being cheap is no longer the primary goal, this idea is about a possibly all-cube-smart-cube system!(Of course, I still want to make it as cheap as possible, but $50 or less probably isn’t possible)

(Sorry @Etotheipi ,I failed to make the cheap smart cube you have hoped for, but someone out there will definitely make it in the near future, probably using a different insight compared to mine!)

Currently discussing the idea with @Lelouch, who is experienced in neural networks.

Woaj deep learning?