A Web-Based Pong Game Against a Multilayer Perceptron Controlled Agent Using Motion Sensing

This study explores the area of integration of the Internet of Things (IoT) with traditional video game paradigms. A web-based Pong clone is developed using the MPU6050 Motion Sensor, which combines a 3-axis gyroscope and accelerometer, allowing real-time gesture-based control. Utilizing the Arduino platform as an intermediary to communicate between the sensor and the game, this system allows players to interact through physical movements instead of traditional input devices. Additionally, an AI-controlled opponent is implemented using a multilayer perceptron to simulate intelligent and human-like behavior to provide a challenging experience for users. This study also reviews the historical landmarks in video games, human computer interface and motion sensors. Furthermore, the latency of the controls is observed to be minimal along with the AI opponent boasting a high accuracy. Moreover, these results contribute to the growing field of interactive gaming interfaces, demonstrating the potential of combining IoT devices and artificial intelligence in web-based applications, with applications in fields ranging from human computer interfaces and accessibility in software platforms.