OpenChess

Hi all, I have started a project to make my own chess sensory board. I am using a BBB (beaglebone black) for detecting the position of the pieces in the chessboard. I want to test it against a UCI engine running on any chess GUI on a desktop PC. The BBB can connect to the PC in many different ways (SSH, telnet...). My question is what would be the easiest way to do this? Maybe make the BBB act as a chess server, or maybe write a sort of UCI engine that I can use on the PC, which connects over SSH with my code on the BBB? I am sure all of this must have already been done but I'm having trouble finding a clear solution, so any pointers will be welcome. Thanks in advance!

The most versatile way to interface with a sensory board is to disguise the driver as a 'pseudo-engine'. You can then use it under any GUI that could run the engine.

the most important question is, what are you using to "sense" the chess pieces?

My idea was to use the principle of the metal detector: put a wire coil (etched on a PCB) under each square, make it part of an LC circuit,and measure the resonance frequency. A metal object in the field of the coil blocks the field lines, and thus alters the inductance, which shifts the frequency. Closed loops around the squares help to confine the field, so that it does not spill over in neighboring squares,and also senses there.

The pieces would be 'metalized' by sticking aluminum foil on their bottom.

I never gottoactually trying this out.

many decades ago I played with the idea of using magnets on the bottom of the chess pieces and reed switches under the squares. i designed a discrete TTL circuit to scan the squares looking for a change in the state of a reed switch. now i would do it with a micro-controller. anyways, i couldn't find good magnets back then and dropped the project. there was also the problem of the magnets wanting to "get together" with neighboring pieces.

the new commercial designs are nice but very pricey. it would be great to find a DIY solution that wouldn't take more time/effort/cost than one of the commercial boards.

finding the right kind of sensor is the key, me thinks.