Robotic arm based game companion

Abstract

Loneliness is a widespread issue that has various mental and physical health effects. Social interactions and connections are essential for curbing loneliness and fostering a sense of belonging. However, for individuals with limited social opportunities, engaging in activities such as games can provide a valuable alternative to ease feelings of isolation. This project explores how a robotic arm game companion can replicate the interactive and strategic nature of human gameplay, offering users a sense of companionship through engaging interactions. This report documents a 32-week journey of developing a robotic arm game companion aimed to reach out to people who experience loneliness, filling the missing human interaction in their lives. By emulating key human capabilities of the brain, hand, eye, ear and mouth, the robotic arm replicates the engaging experience of playing with a human partner. The system integrated tree search algorithms for strategic decision-making, allowing it to intelligently counter a player’s moves. Servo arm control logic enabled precise object manipulation, allowing the robotic arm to pick up and move game pieces smoothly. Various computer vision techniques were used for detecting game objects and reacting to game moves, aiding informed decisions during gameplay. Speech recognition allowed for verbal communication, enhancing the sense of companionship. This report also explores the challenges encountered during development and strategies employed to overcome them, ensuring product quality while ensuring the necessary interactions are offered. The Yahboom Dofbot with Raspberry Pi was used for the development of the game companion.