Project Description

This Dissertation project showcases an intersection between gaming and artificial intelligence, specifically focusing on the creation of a dynamic and competitive racing game enriched by the sophistication of machine learning. Leveraging the capabilities of Unity and its ML-Agents framework, the project pioneers in training AI racers through reinforcement learning, particularly the Proximal Policy Optimization (PPO) algorithm, to challenge human players. The AI agents evolve through a meticulously designed multi-stage training process, gaining proficiency to adeptly navigate complex racetracks and tactically outmaneuver opponents. This project not only demonstrates the potential of AI to elevate gaming experiences but also serves as an exemplary model of utilizing Unity’s advanced tools to craft intelligent behaviors.