The PongBot was a project for the 2016 Ohio State Makeathon, a competition where teams had 24 hours to create a unique project utilizing software and hardware. After our original project derailed, the Pongbot was produced and coded in about 12 hours. The novelty and execution of the idea attracted a lot of attention from judges and other competitors.
On the hardware side, the robot was constructed with a wooden frame and actuated by three servos. Two servos were used to controlled the rotation of the robot perpendicular to the ground, and the angle of the launching arm relative to the ground. The third servo was used to create a consistent force on the spoon for launch. An Arduino was used to control the angle of each servo.
In my opinion, computer vision was the most interesting aspect of this project. A raspberry pi running OpenCV used a camera to determine the distance and angle of each remaining cup on the table. The Raspberry Pi selected the cup closest to the camera as the next target for Pongbot. This information was sent to the Arduino, which selected the proper servo angles needed to land a ball in the cup.
Toward the end of the event, Pongbot attracted a lot of attention from many other competitors and judges. Several tried their hands at beating the robot in a friendly game, and many could not! This was a fun project, and greatly increased my knowledge and appreciation of computer vision applications.