In partnership with a classmate for our Introduction to Robotics course, we programmed two robotic arms in MATLAB to play Pong, the 1972 Atari video game, but in real life. Both arms shared the same video feed, where a camera tracked a red ball using color segmentation. From just the last two frames of motion, the system predicted the ball’s trajectory—including when it bounced off the walls—and sent commands to each arm. We implemented inverse kinematics from scratch to map those predictions into smooth arm movements. The result was two robotic arms rallying back and forth, anticipating each other’s moves, and even scoring points—essentially playing a continuous match of Pong without human input.