An idea for a "physical interface for OGS" - Feedback required

Hi,

I am technician and Go enthusiast and wanted to get some feedback on an idea. I am not sure if only me is missing something like this or if others have the same need - hence this post.

I love OGS but I do not love playing Go on my Laptop so much. Hence, wouldn’t it be nice to play on a real nice wooden board with real stones and still play on OGS? Puzzled?

Let me explain:
What I am thinking of is a real, wooden Go Board that has very tiny LEDs on each crossing to show to you where your (remote OGS) opponent placed a stone and sensors that recognize that you placed a stone. All of this is connected to OGS so that you can play a real haptic Go game without the need to have the physical opponent sitting right next to you.
The heart of this solution would be a Raspberry Pi Mini PC which actually costs ~$30 and has all you need to implement the internet part as well as the control of the LEDs and sensors.

Question: would that be something you would like to play on or is it just a weird wine idea?

Looking forward to your feedback
Guido

13 Likes

I think it could be quite a nice thing.

I wouldn’t call it “Go Board Computer” because that brings a sense of AI play vs computer etc.

Maybe “physical interface for OGS”?

I presume it would have an LCD mounted styishly on one side for the find-a-game part of things - rendering those OGS pages as HTML, and only rendering the Game page itself via LED sensors…

Wouldn’t you also need 361 photoreceptors so “the board” knows where you played?

Edit: Somehow I skipped the part in your text where you mentioned ‘sensors’. In that case, myea.

Not for me, sadly, because if I have to place stones for two people in a blitz game… that’s bound to fail. Probably not that interesting for correspondence players, either. Leaves us with the live crowd.

I think you’d be better off just producing a mat people can put on their real boards. That way your production is cheaper, faster, lighter (shipment!) and generally more versatile. You can wire my consultancy fee to…

3 Likes

I knew a guy who was working on something along these lines several years ago, and we kicked around several ideas.

Individual photoreceptors seemed like the most natural solution, although without some cleverness that may be a lot of inputs for a stock Raspberry Pi. The market for such a board would be small-ish, but I think it would be cool to have one.

Of course there’s the added challenge of having to also place the opponent’s stones, but what’s to be done?

Yes, that is true, it will need a lot of receptors. Hence, I do not expect to hit a big market with it, it is more a hobby project. If someone would be willing to pay (quite some) money for something like this, I would love to produce them on a very small scale without a profit-generating mindset. I am not looking for a business here but rather wanted to check if this is something that already exists somewhere or if there are other ways to solve the issue.
A mat would be a technical solution but lacks the esthetic aspects that I consider to be important to beat the computer-only-experience.

Guido

2 Likes
  1. Purchase $3000 kaya board.
  2. Drill holes.
  3. ???
  4. Profit.
4 Likes

I don’t think that a Raspberry would have issues controlling a 19x19 grid of sensors and LEDs, do you?

Having to place the opponent’s stone as well might even lead to a better play in the end. It requires you to pay more attention to the opponent’s plan.

Guido

1 Like

My wooden board was ~$60 and looks very nice :slight_smile:
As I said, this is not about making profit but making the computer-Go-world more haptic.

The Pi should have no trouble with the number of sensors. It’s just a two-dimensional array.

But we all watched Hikaru be grumpy about having to place Sai’s stones for a few years.

2 Likes

Exactly. As thew focus is on esthetic the board would have to be very clean (expect for the sensors that need to be as invisible as possible). Everything else could be controlled via a mobile app or a small 3D-printed box with simple display that allows you to do basic controls like “find an opponent”, “pass”, “accept score”, …

I don’t think it would be a big success because it isn’t useful for many players (fast live and correspondence, as said before) but it could be anyway a cool piece of hardware. :grin:

Could it also have a Bluetooth interface for use with mobile/pc?

It could but this would make it even more complex :thinking:

1 Like

Related video:

5 Likes

All kidding aside, I would love to have a physical board as my primary OGS interface. But it would be cost-prohibitive, more effort, and a cat magnet.

1 Like

This is already done, but without sensors, or leds, but camera and image recognition, and a bot for placing the stones, solution that solve the issue pointed by few, of placing opponent stones.

Anyway, it was not done for OGS or any other server, but for an Go program. For sure can be adapted for OGS. Here are the details. http://blog.springwald.de/blog/post/igobot-a-go-game-playing-robot-on-raspberry-pi-using-opencv-and-gnugo

And here is a video for it https://youtu.be/1TQPHkmvTog

And a demonstration game. https://youtu.be/UeOAhcMHR-U. Nicely done, except that the stone could have been refilled while the human is still thinking, not while the human waited after he placed his stone.

6 Likes

(Almost) No one would put a contraption like that in their room just to play OGS games on a board.

OP’s suggestion is at least subtle enough to make sense.

2 Likes

Something about this doesn’t appeal to me. If I play on the computer, I want a computer interface (or mobile, which is a similar concept). The interface is already easy enough to deal with, and there’s minimal setup. If I play on a real board then I want simplicity and quality. The aesthetics of a clean brown board with white and black stones is attractive to me. Adding a robot or blinking lights isn’t something I’d ever do.

The use case is ‘people who want to play on a real board but only (or mainly) have internet people to play with’.

I would not say nobody, The guy who made, for sure already did :wink: And at least one would can be found. For sure, yes me neither, But the camera and image recognition for sure I would use. And for pointing the opponents move I would use a laser pointer, It may be technical more complex to implement, but maybe cheaper than 361 diodes.

I hate placing stones while replaying games, and writing moves of my own games so I invented myself such contraptions. It was only theoretical. No prototype produced. One is even uglier than this. It consist of an array of mechanisms, similar of those from a ball pen. But not for sliding out or in the point of the pen. in and out would be a semi sphere white, one black and a planar one so, the board looks normal, with some slight circles around the intersections. When you press the intersection the mechanism will slide inward and turn to reveal the black side. press again and reveal the white side. That is, white has to work twice as hard. A simpler version is that the three surfaces are placed on an axle, you press on a side it will rotate 120 degrees and reveal black, press the other side, reveal white. And at all intersection could be some connectors that can send to a computer the moves, to be saved in a sgf. This is not economically feasible, but I could not stop myself inventing it. Luckily the camera recognition was produced. I thought of it too, of course.

And my obsession for this thing was satisfied once more when I found about such a goban that was actually produced in Japan. I wrote an article about it.

I copy the pictures here.

and a an actual photo of it.

It was similar in idea, an array of leds and sensors under surface. It was produced before internet era, to connect players via phone and modems.

Guido, you may be lucky to find one of these and adapt it

4 Likes

I agree. It is a challenge to keep up the aesthetics of a real board. However, if there a really small sensors and LED it might be possible to hide them pretty well. You would never see a blinking light only if you have to place an opponent stone. So, this should be rather non-intrusive.
The robot solution is interesting but I agree that I would not place such a thing into my house. I also think it kind of destroys the clean board feeling that I am envisioning.
Guido