Automated Pacman

AI: State machine & A* pathfinding

To demonstrate my understanding of state machines in AI, I chose to create an automated Pacman game as my project for the AI in video game class. I implemented two types of finite state machines to govern the behaviors of both the player and the enemies. Additionally, to enable efficient movement for all characters on the map, I integrated the A* pathfinding algorithm using a node system. This algorithm calculates optimal routes, considering walls as obstacles, allowing characters to navigate the environment intelligently. The combination of state machines and A* pathfinding enhances the gameplay experience, providing autonomous and strategic behaviors for the player and enemies, as well as smooth movement across the game world.

Behaviour optimizations

In order to enhance the natural and intelligent gameplay of the characters, I implemented behavior optimizations. This involved incorporating distinct targets and a line of sight mechanic using raycasts. By utilizing these techniques, the enemies were able to prioritize high-priority objects, such as pills and pellets, while also reacting strategically based on their current state. For instance, the player character would target pills and pellets while evading enemies when in the wandering state. However, after consuming a power pellet, the player character would switch to a chasing behavior, actively pursuing and attacking the enemies. This implementation adds depth and complexity to the gameplay, creating a more realistic gamplay style.