EMBEDDING ARTIFICIAL INTELLIGENCE IN GAMES NPC AND EFFECTS ON EMOTIONAL HEALTH

Abstract

Artificial Intelligence (AI) has transformed the role of Non-Playable Characters (NPCs) in video games, enhancing their realism, interaction, and narrative depth. In this study we have discussed the most used Artificial Intelligence techniques in games NPCs like pathfinding algorithms, decision and behavior trees, emotional modelling and adaptive difficulties and concluded the best applicable techniques. The main advantage is to determine and separate the most effective techniques based on time and space complexity from non-effective techniques and to tell how they one is better than the other and in what way. This is achieved by performing comparisons and identifying differences between difference techniques along with the explanations and examples. To validate the accuracy of the proposed method, we compared Dijkstra’s algorithm and A* algorithm with the same problem statement and conducted 2 tests between them.  Both time A* algorithm outperformed Dijkstra’s algorithm in time complexity and 1 time in space complexity. It was found that the proposed method was accurate and determined the best techniques among others. This method is useful in making an open world realistic game where NPCs are roaming around the game world or driving vehicles in the game as it makes the game more realistic with realistic NPCs, which can be achieved by using the techniques proposed in this study.