AI Based Batting Simulation Machine

Abstract

The introduction of AI has transformed the world in various aspects such as sports training. This thesis represents the evolution of bowlers' training in cricket. This system integrates real-time image processing, machine learning techniques and electromechanical control to mimic human-like batting response. The Raspberry Pi model B acts as the brain, detecting and tracking the ball using a Pi Camera 2 Rev 1.3, leading to trajectory processing , consequently activating both TD-8125MG servo motor and Nema 23 stepper motors to manipulate the mechanical bat. The primary objective of the system-to assist bowlers without reliance on human batsmen, was achieved through the creation of an integrated and effective batting machine. The project experienced setbacks in aspects such as synchronization, power management and latency minimization. These aspects were dealt with through prototyping as well as software optimization. Though the system fnally achieved its goal, several limitations pertaining to ball detection and dynamic obstacle handling were observed. The AI batting machine highlights untapped potential for cricket coaching, academics, schools and individual players. Future versions may reinforce integration of new components such as high speed cameras, advanced machine learning and wireless communication. This research highlights how training methodologies in sports can be updated through integration of embedded systems, machine learning and robotics.