Team Turn Down for Watt
The goal of our project is to produce an air hockey opponent with artificial intelligence that can competitively play against any user. Recently, older tabletop games have been remodeled and engineered to include robotics. This allows the user to feel as if they are playing against another human player which eliminates the dependency of game play on the presence of another person.
Our design approach centers on a vector prediction algorithm that enables the AHA! to predict the location of the puck and move its mallet to defend its goal or strike the puck back at the user. The puck is monitored via a camera mounted at a birds-eye-view of the air hockey table which feeds images to a microcontroller. The AHA! mallet is attached to a 3D printed and laser cut rail system driven by DC motors and uses encoders to establish an XY coordinate system on the table. Users can interact with the system through an LCD touch screen interface. Options are given to calibrate the system, choose a difficulty setting and start the game. During game play, the score is displayed for the user to see and the user also has the option to terminate the game prematurely.
Upon completion of the project, we foresee one area of advancement upon our system. This includes a user feedback module that monitors player movements and defense/offense play to report areas of improvement that the player can make. This would require a much longer time frame, but is a feature that consumers tend to expect from modern games.