Connect Four Physical AI

  • Connect Four is a classic strategy game in which players take turns dropping colored discs into a grid, with the goal of getting four of their own color in a row either horizontally, vertically, or diagonally. In our project, we took this classic game and turned it into a physical AI by using 3D printed parts to create a physical Connect Four board, and using image recognition technology to scan the board and determine the state of the game.

    To make the board interactive and able to play against a human player, we implemented the minimax algorithm, which is a common AI technique used in decision-making and game-playing. This algorithm allows the board to "think" several moves ahead and choose the best possible move based on the current state of the game.

    To physically make the moves on the board, we used Arduino servo motors, which are small motors that can rotate to a specific angle and hold that position. By using these motors, the board can drop the colored discs into the grid and make its own moves.

    Overall, our project combines physical hardware, image recognition technology, and AI algorithms to create a unique and interactive version of the classic game Connect Four.
  • Andrew Stekar


    University of Waterloo

  • Callum Thompson


    Western University

  • Andrew Scott


    University of Windsor