Prototype Fuzzy Logic AI

Use WASD - Move

Use Left click - Shoot

Overview

Implemented a Fuzzy Logic system to enhance AI and gameplay mechanics. This approach was designed to facilitate smooth state transitions and optimize reaction times when handling multiple variables. It was primarily integrated into the Enemy AI to dynamically manage behaviors and transitions based on player interactions.

Main Goal

The primary objective of this prototype was to evaluate the system's versatility. It demonstrates that smooth transitions can be achieved using significantly less code, completely bypassing the need for traditional Finite State Machines (FSMs).

Pros & Cons

• Pros: A Fuzzy Logic system is highly effective for state transitions, offering precise control across diverse variables. It reduces the required codebase and performs efficiently, even with a large number of entities. Additionally, the logic is highly intuitive, as it converts exact numerical inputs (such as specific speeds or distances) into linguistic variables and fuzzy sets using membership functions.

• Cons: The process of designing the logic curves can be somewhat confusing. Furthermore, setting up the membership functions requires strict attention to naming conventions to avoid detection errors.