> Functionality and Features
> The Pi Board is an advanced automated chess system powered by a Raspberry Pi, utilizing an XY stepper motor mechanism and magnets to move chess pieces seamlessly across the board. The development process involved several key stages, including precise calibration of stepper motor coordinates, calculating the weight of each piece for accurate handling, integrating a robust chess engine, and optimizing piece-grabbing strategies and movement detection. Special attention was given to selecting the most efficient algorithm to minimize the stepper motors' power consumption.
Is there a reason for the marketing speech? I'm assuming most people interested in this would rather read engineering speech, like so:
> The Pi Board, as the name suggests, uses a Raspberry Pi under the hood to calculate engine moves from <Stockfish? Leela Zero?>, and move the pieces with a series of stepper motors and magnets. We spent a significant amount of effort minimizing power consumption, including weighing the individual pieces to get more efficient grabbing and moving motions for each one.
Probably an LLM-assisted blog post.
It must be. The part about minimizing the stepper motor power consumption is nonsensical. Steppers use the same current whether moving or stationary.
You can absolutely adjust the current for a stepper motor. Lower currents use less power, but have less torque, leading to an increased chance of missed steps. In a system like this, the motor torque becomes a linear force to move a belt — and the force required to move the belt depends on if a piece is moving with it, and which piece. It’s perfectly reasonable to bump up motor current when moving a piece, and bump it down when moving just the belt to grab the next piece.
> Steppers use the same current whether moving or stationary.
You can turn off the power when you're not moving it, assuming it doesn't need to hold anything in its place (like in this application).
E.g. https://www.allegromicro.com/-/media/files/datasheets/a5984-... describes how its current control in general works for this stepper motor driver in page 9 (certainly not using the same current always) and in page 12 how to use its ENABLE signal to control FETs and SLEEP to further reduce power usage.
Only the most naive stepper driver will drive a stationary motor at full current
That's an immediate disqualifier for everything (for me) unfortunately.
I would suggest keeping an open mind. I use LLMs to check, my writing and if could be could improved. English is not first language and I am not confident enough to know if what I wrote comes across professionally or not. May be the author of the blog post has the same issue.
English is my fourth language and I much prefer reading something that is imperfect but written by a human to something auto-generated. That feels as if we someone would say "I'm busy, talk to my agent instead."
This comes across a little aggressive and braggy. Perhaps a better way to write it would be:
> I much prefer reading something that is imperfect but written by a human to something auto-generated. That feels as if we someone would say "I'm busy, talk to my agent instead."
Condescending to someone who speaks your language as a fourth language about their tone comes off as extremely rude and superior. A better way to write this would be:
""
I'm not certain that this blog post is the original. The bottom of the website has contact information for the author, and the hyperlinks for the email and phone number link to different emails and numbers than are listed on the page.
Section 2 has the words "Mars Science Laboratory" (the official name of the Curiosity Rover) printed vertically to the right for no discernible reason.
It's almost as if this were a patchwork of content from multiple sources, maybe glued together by an LLM.
It's hosted on a free hosting platform with no chain of custody. The author is only known as "Tamerlan". Why wouldn't an engineer have their own blog for a project like this?
Yeah it really does read like something spat out by ChatGPT. They always have to include the word “key” for some reason.
Indeed, I see the word "key" in titles of many click bait LLM generated articles. Probably it is creating a self feedback loop now.
I wish they added some LLM-assisted CSS & Div tags because this layout is atrocious.
Hate to be a spoil sport but if when pieces move they nudge other pieces out of the way and then those nudged pieces have to be put back by hand, then there’s still work left to be done. Same goes for capturing - the captured piece should walk itself off the board.
My first thought was "those nudged pieces can surely be put back in place automatically using the same exact mechanism!".
I was also thinking that each piece could have a magnet at the base. And when a piece is captured, the underlying gantry could flip the polarity with an electromagnet and sorta launch the captured piece off the board.
Although it seemed like the author had other issues with using an electromagnet here (i.e. overheating / switching polarity).
For a hobby project seems like good progress.
Polishing the last 20% of functionalities will take him a lifetime. Maybe other chess/robotic enthusiast could join him to accelerate some border cases.
I don't see this as "labor having for the human" but rather as a way to immerse yourself into the fantasy that you're playing another being.
Yeah the videos were far from impressive...
On the other hand, it's certainly more impressive than almost anything I've put together in the past five years.
This brings back memories of Regium. A kickstarter scam 5 years ago that used realistic 3d animation videos of an automated chessboard to trick people out of their cash.
chess.com happily took regium's money to advertise their scam to their audience.
It eventually got kicked off kickstarter, and then kicked off some other kickstarter clones before self hosting their own kickstarter clone website. That was a wide ride. searching youtube for "regium chess board" will get a few hits.
Pi board looks like a fun project that I'm sure will be refined over the years.
Not sure if I missed it, but I didn't see any mention of how player moves are detected. Is there a camera and CV? It's always seemed like a fiddly problem to solve with electronics as you end up needing a matrix of sensors so there's a lot of wiring.
I’ve always wanted to try making a smart chess board (with no moving parts; merely detecting moves rather than making them).
I’ve thought of:
- RFID. Have 64 antennas and multiplex them to detect which piece is on which square (idk much about RF so this felt tough)
- Vision with a fiduciary mark under each piece, and an acrylic board
- Hall effect sensors, where instead of knowing which piece is which, it instead assumes the normal starting position and pays attention to which square was picked up from and which square was placed onto to infer which piece moved.
I think with any of these approaches it’d be fun to make a tiny, single-PCB board.
> RFID. Have 64 antennas and multiplex them to detect which piece is on which square (idk much about RF so this felt tough)
The professional-level boards by DGT use RFID and retail for about $500.
I looked into building a competitor some time ago. 64 RFID antennas alone would have eaten up that budget. I believe they do something smarter like having 8 antennas and arbitrating the signals. They have some patents in this area.
I've seen Hall effect and barcode-based systems too. They've always been a bit less reliable than DGT. Actually DGT is not all that reliable: if you are broadcasting a 20-player tournament you will need to manually update the broadcast around once per round. However I think the position detecting (hardware side) is solid and they could do with improving the move detection in software.
None of this is meant to deter you from building this as a hobby project! I think all the approaches would be fun to try.
DGT uses a (patented, expired [1]) resonant LC circuit, not RFID.
Extremely interesting, and kind of ingenious. Thank you.
I was trying to do something like this a while back, our approach was to have a different color underneath each piece, with an elaborate setup to get the colors reflected into a camera, but I could never get the color detection working reliably with the way we were doing it. It was a fun project though, there's got to be some easy way to detect moves and get a cheap-ish internet-enabled board.
Yeah there are some trainer chess boards that use hall sensors to track piece movements. But I think there is a possibility to actually encode pieces with different magnetic field strengths and flip them for each player. That way you can just do stateless reads and you'll always get the correct readout, plus you can recover from illegal states.
I did a project [0] a few years back that did this absolute encoding for senet, since there is only one type of figure and two players so just flipping the magnetic field worked really well once calibrated. I still need to make a proper writeup/video on that thing one day...
I don't think 64 RFID antennas would be that bad - you can etch them onto a PCB. That would be a pretty large PCB, I guess, but you could segment it if necessary.
Put an ultrasonic emitter and an accelerometer in each piece. When a piece completes a move emit an ultrasonic pulse pattern unique to the piece. Pick that up with 3 ultrasonic microphones places around the board and use the time differences between when the pulses arrive at the 3 microphones to find the location of the piece.
Maybe use 6 different frequencies (one for the white King, one for white pawns, and one for the rest of white, and similar for black) to make it easier to handle moves that affect more than one piece.
The moves that involve more than one piece are captures (one piece of each color), castling (one King and one Rook), and pawn promotion (one pawn, one piece of the same color that the pawn promotes to, and possibly one piece of the opposite color if the pawn captures during the promotion).
Yeah, the transparent board with fiduciary marks and a camera was the road I was going down at one point.
There's also a really neat solution that involves a matrix of wires, one for each row and column, and then running a current through the wires. The pieces then each interact with the current in a unique way producing a signal. You then do some clever stuff to figure out all the pieces that are on a given wire. Cheaper than RFID, and you don't need a sensor for each square and can do some multiplexing. I think it possibly uses magnets in the pieces? You can tell electronics is my strong point!
The movement of the knights is so creepy somehow. Lots of little movements and rotations. And then they just push a pawn out of the way and the human has to put it back.
The castling at 2:10 was also pretty lazy, but in fairness some humans also leave the board like this after moving :P
I couldn't get over how much background noise they had in the video. If that's where they work I would go crazy.
I wonder if using two "arms" under the board would result in better movement - you'd need two motors to control the angle of each arm, and some way of extending the arms, but then you'd be able to use one arm to move any required pieces out of the way, while the other arm moves the intended piece around. You'd use a polar coordinate system, instead of a cartesian one, to keep track of where the arms are and where they need to move.
Plus, you'd be able to move the pieces off the board when they're captured.
I think you are describing a 5 bar linkage or parallel scara robot.
Not quite. I'm imagining two independent arms, probably placed at opposite corners. They'd have one servo for rotation along a 90 degree angle, and another mechanism for extending or retracting it. I can't really draw something at the moment to illustrate what I mean.
(But practically, yeah, what you linked is probably what would be required, otherwise I don't know if it could support it own weight.)
I absolutely remember encountering one of these at a friend's house. I clearly remember the detail about moving the pieces out of the way as needed to make moves. I do not remember the noise of the xy mechanism.
The noise is likely the result of the recording method. I don't remember it being that loud at all.
Hey guys, I am Tamerlan the creator of this chess board. To answer some questions about the background noise its since I am working in my garage. Currently as a high schooler I don't have too much time to make improvements to the board such as having the piece automatically removed when captured, and an additional robotic arm underneath to move the pieces out of the way. I just decided to use an approach to move the pieces through the middle of the squares. Hopefully in the near future I will keep improving the pi board. Engineering has been my hobby for a while, since I like playing chess I decided to connect them two. I am not much of a web dev so the website is not the best, I made it to show the main idea of the concept, in the future I will improve the website as well and link it to all of my other projects I've built in the recent years. I would like to discuss with all of you more about the board, if you have any questions just shoot me an email at thstudios0708@gmail.com and we will get in touch! Thanks for all the support.
It really, really struggles to move the pieces. More development is needed.
Really neat... very impressive work so far, so inspiring!
It's kinda funny to see just HOW MANY of these automated chess things exist... it's like people really do want the physical experience while still playing "with" someone. As I'm guessing OP will discover, the devil with this is in the details, and getting all of the last mile mechanics to work quite right is probably a huge pain to truly be able to seamlessly play against the opponent with the pieces moving automatically.
Surely the "actual" answer to this is some robotic arm with good enough camera work? All the things that make this hard -- the pieces not being placed perfectly, the knight jumping / or pieces that are being captured in really crowded places gets theoretically solvable with that approach. Of course then you have a huge issue with the engineering precision on the grabber and camera lol.
Cool project!
Could a hefty reverse magnetic pulse launch captured pieces off the board? And maybe with some precision by controlling the offset and power?
Battle Chess 2024: more exciting than the original. The human player has to dodge the assault of their electronic opponent!
Sometimes an interesting idea is best left unimplemented.
Man, I remember playing Battle Chess back in the late ’80s. We’d go out of our way to see what all the different piece vs piece conquests looked like. I’d love to have that on the iPad for my chess-addicted son.
I was thinking shooting the piece into a basket or something, but aiming for the player's forehead sounds fun.
Maybe slimmer pieces would help prevent collisions during movement? (It is a pretty obvious suggestion, so likely there is a reason the author didn't take this approach. Is the lower bound of the piece's base determined by electromagnet strength?)
The pieces look to be well below normal size for that board. At normal size it would be very difficult.
To be USFC/FIDE legal, the board has to size with the pieces. Even though it's not likely to be used in a tournament, it would just look odd being sized differently.
The fact that we have tinkerers building their own working versions of a well-known 1980s Brookstone novelty gadget makes me smile. Reminds me of the guys who built their own Segways back when we thought Segways were things people actually wanted.
I watched through all of that game play footage just to see if a castle would happen, and it did happen at the end, with the same bumpy movement that ran throughout the game.
The most interesting bit was probably near 1:50 where the pawn moved sideways.
>the pawn moved sideways
Holy hell!
Love this!
Also for those interested in auto-moving chess boards, see the Go Chess project:
https://www.indiegogo.com/projects/gochess-most-powerful-che...
It takes a different approach: 1-4 "robots" under the board moving pieces around - with the goal of being able to reset the board or setup a position super-quick!
The issue with these "ghost chess" implementations are always the same things
1. How do captured pieces get removed? 2. How to handle pieces that move through other pieces (knights, castling)? 3. How to promote a pawn?
I have never seen these solved in something that is not awkward and neither seems to be the case here.
1. I see what you mean. Without a dual gantry it'd first move the taken piece off of the board and abandon it, then move the capturing piece. The meantime'd be awkward. Perhaps he could move the piece to the corner of the current square, move the capturing piece in, then move the captured piece off the board. That'd be slow and tedious to watch. So. No great solution, which was your point I think.
2. Personally I thought this implementation demonstrated this pretty reasonably both with castling and with moving the knight. He kinda "cheats" in that his squares are pretty large compared to the pieces, so the jumping piece can just travel /between/ ranks and files. Seems pretty nice to me.
I've seen these kinds of chessboard automaton systems before, and they commonly struggle with complex movements, such as castling pieces, capturing, and the movement of knights.
A lot of them can't handle "contingent movements" where you need to displace piece X but in order to do so, you first have to displace piece Y, and then rewind the stack to restore the original pieces positions.
The second video where they demonstrate the actual gameplay footage exhibits some of these problems with this board, e.g. pieces that are displaced don't seem to re-center automatically, etc.
A real mechanical turk. Needs to be able to handle captures / castling too. Maybe I'm wrong but I don't see too much practical purpose for this, but as a piece of the scenery it could be very cool.
I love the board has no led when X the opponent. You can add LLMs as your opponent. So you can play with your dead grandpa, hitler, aristoteles while talking about everything. I'll buy or made one ! :D
Nice. Perhaps would be even better with a coil array and no moving parts.
So if there are challenges in moving the pieces from below, then why not have a robotic arm that reaches down and grabs and moves pieces? A bit of computer vision might help with accuracy.
Because that's even harder.
I dunno. We can pick up bowling pins pretty easily. Why not chess pieces.
We are also moving away from traditional bowling pins because the mechanisms are too expensive and hard to maintain, and moving towards pins that are attached to strings, to cheaply reset
Bowling pinsetters are rather infamously dangerous machines and operate with all kinds of guards to keep meaty appendages out.
For one, chess pieces vary in shape and height.
Reminds me of squareoffnow[.]com. I've been wanting to get two boards for a while now and do remote play with friends.
The author mentioned at the beginning of this post that they were inspired by SquareOff board demonstrations.
hi, your website has spelling issues and also the contact text doesn't match up with the links?
"knight Movemetns and Casstle?"
Really cool!
Side note: there are some typos in the last header :P
Switching a pawn to a queen is going to require some creativity :)
How does it move the knight?
Interesting approach! I like how moves with pieces in the way were solved for the most part. A bit disappointing that the taken pieces have to be manually removed -- one way to handle that could have been to extend the range of the "head" to the margin of the board, and have it drag off the captured piece to that area before moving the capturing piece there.
I wonder if another mechanism altogether could work a bit better -- a matrix of electromagnets embedded under the board, with a resolution of say 1/4", so that when deployed in sequence, they could move the pieces from one magnetic field to another and another, without any actual moving parts involved.
> I wonder if another mechanism altogether could work a bit better -- a matrix of electromagnets embedded under the board, with a resolution of say 1/4", so that when deployed in sequence, they could move the pieces from one magnetic field to another and another, without any actual moving parts involved.
A standard chessboard has squares of 2-2.5”, so this would require at least 4096 electromagnets.
There was a kickstarter [1] for such a design a couple years ago. The conclusion in the chess community was that the kickstarter was a fraud [2] (with faked videos) and such a board would not be commercially feasible.
[1] https://www.kickstarter.com/projects/soontech/regium-automat...
[2] https://lichess.org/@/lichess/blog/regium-extraordinary-clai...
I was just thinking catapult taken pieces off the board with a reversed strong magnetic jolt, so they just shoot off like a rocket.
1. e4 e5 2. d4 Bc5??
Call the exorcist...
Truly a mechanical turk
That was still operated 20 years ago: https://www.hnf.de/en/permanent-exhibition/exhibition-areas/...