Nearly three decades have passed since chess legend Garry Kasparov was famously defeated by IBM’s Deep Blue, marking a historic moment as the first time a reigning world chess champion lost to a computer in a match. Since then, chess engines have evolved exponentially-today, even basic smartphone apps can challenge and sometimes outplay grandmasters. Despite these leaps in AI, the physical act of moving chess pieces has remained a human task. That is now evolving.

Recently, Joshua Stanley Robotics, a maker and YouTuber, unveiled his innovative DIY project: a physical chessboard capable of recognizing human moves and autonomously moving its own pieces. Like several earlier self-operating chessboards, Stanley’s design leverages magnetic technology. Each chess piece was 3D printed with a hollow interior to embed a magnet at its base. The board itself is constructed from a printed circuit board (PCB) embedded with magnetic sensors that detect the position of each piece in real time.

Movement is achieved through a motorized system beneath the board that maneuvers an electromagnet along the underside. When activated, this electromagnet attracts the magnet inside a piece, dragging it smoothly across the board to its new square before disengaging.

The chess engine powering the decision-making is the widely acclaimed open-source Stockfish, which allows Stanley to dynamically adjust the AI’s skill level. Interestingly, Stanley admits he is not an avid chess player and prefers to keep it that way.

“Rather than investing time in learning or practicing chess, I decided to build a chess robot that can defeat me so decisively that I lose interest in playing,” Stanley explains in his project video.

Designing an Autonomous Chessboard: Challenges and Solutions

Stanley approached his project by addressing three core challenges: detecting human moves, computing the AI’s response, and physically moving the chess pieces. While the first two are relatively straightforward in software, translating them into a tangible board posed unique difficulties. Embedding magnets inside 3D-printed pieces allowed the system to identify and manipulate them effectively. To differentiate sides, Stanley assigned opposite magnetic polarities to black and white pieces, enhancing the board’s ability to track each move accurately.

Initially, Stanley attempted to develop the chess engine code himself but soon realized it was beyond his expertise. He opted to integrate Stockfish, a powerful open-source chess engine, to handle gameplay logic. To bridge the physical board and the digital engine, he created a Python script that converts sensor data into a format Stockfish can interpret and translates the engine’s moves back into physical commands.

Before settling on magnets, Stanley experimented with a robotic arm designed to pick up and place pieces. However, the arm lacked the precision and reliability needed for consistent gameplay. The magnet-based system proved simpler, lighter, and more portable, though it does have some limitations. For example, knight moves-which involve jumping over other pieces-can cause collisions, occasionally toppling pieces that must then be reset manually. Additionally, captured pieces need to be removed by the player.

Despite these minor drawbacks, Stanley considers his creation fully functional and enjoyable to play against.

“I’m thrilled with how this project turned out,” Stanley shares. “The subtle hum of the motors and the hidden electromagnet movement add an exciting layer of suspense to every move.”

Stanley’s project is not the first self-moving chessboard to emerge. Several commercial models already exist, many employing similar magnetic technology. For instance, the Miko-Chess Grand is a tournament-sized wooden board that uses magnets to move pieces and retails for around $497.

Another notable example is the Phantom chessboard, which also uses magnets but integrates with online platforms like Chess.com. This allows players to compete against remote opponents, with the board physically replicating their moves in near real-time.

Compared to these polished commercial products, Stanley’s board is more minimalist and experimental. For him, the project was less about mass-market appeal and more about embracing a new technical challenge.

“This project was a fantastic learning experience,” Stanley reflects. “It gave me a great reason to dive into Python programming, which was an added bonus.”