In the ever-evolving world of video game design, artificial intelligence (AI) stands at the forefront of creating more immersive and challenging experiences. As games strive for realism and engagement, the sophistication of enemy AI becomes crucial. This research delves into the development of advanced AI tactics for enemies in shooter games, aiming to transform how players interact with and perceive these digital adversaries.
Imagine playing a shooter game where enemies don't just charge mindlessly but instead use intelligent tactics to outmaneuver you, cooperate, and adapt to your strategies. Our project seeks to achieve precisely that by enhancing enemy AI to exhibit behaviors such as dynamic cover selection, flanking, and coordinated group tactics. Inspired by the groundbreaking AI systems in games like "The Last of Us," we aim to push the boundaries of what is possible in game AI, offering both developers and players a new level of depth and excitement in gameplay.
Our research is not just about creating more challenging enemies but about making these interactions feel real and consequential. By incorporating advanced AI tactics, we can significantly enhance the immersion and replayability of games, ensuring that each encounter feels unique and engaging. This report outlines our journey, methodologies, and findings, providing valuable insights for game developers looking to elevate their AI systems and deliver unforgettable gaming experiences.
The primary goal of our research is to explore and implement advanced AI tactics for enemies in shooter games. By doing so, we aim to:
Our research addresses common complaints in games like "Battlefield V" and "Halo 5: Guardians," where the AI can sometimes be predictable and lacks depth. By enhancing enemy AI to exhibit sophisticated tactics such as flanking, coordinated attacks, and strategic use of cover, we aim to create a more engaging experience that requires players to think strategically and adapt dynamically.
Our research builds on existing AI techniques and introduces new mechanisms for dynamic and adaptive enemy behavior. The following sections outline the key components of our methodology:
One of the most crucial aspects of advanced enemy AI is the ability to dynamically select the most optimal cover based on various factors such as player position, visibility, and distance. This section outlines how we achieve dynamic cover selection using a combination of vector mathematics and pathfinding algorithms.
Steps Involved:
The game will pre-process the level geometry and compute all corner positions of a wall, each corner will then offset a small distance to get 3 cover points (referred to as “posts”) based on the direction of each corner.
Upon alert, the AI agent first searches all corners around themself (constrained by a set view radius) and saves them into memory.