Developing video games with elementary adaptive artificial intelligence in unity: An intelligent systems approach

Abstract

Video games have increasingly demonstrated a great deal of audiovisual realism, in par with the massive performance improvement of computer systems. At the same time, their Artificial Intelligence (AI) component falls short in terms of realism because it is usually based on non-adaptive methods. Adaptive AI mechanisms can help increase video game realism allowing the game to adapt in real-time to the game progress and the user behavior. Following a short overview of the progress of AI in video games in the past years, this paper highlights the creation of modern video games with basic and elementary adaptive game AI using the Unity game development framework. Particular emphasis is on the details of the AI component. First, a shooter game with basic AI is created. Finally, an action-adventure video game is created featuring elementary case-based adaptive AI. The objective in this game is to create enemies which are able to perceive changes in the environment and adapt their strategies accordingly. Proposed AI practices can migrate into relevant real world applications, such as video surveillance and intrusion detection systems, mission critical autonomous networked patrolling and/or save and rescue robots, vision and hearing assistive applications, intelligent video and behavioral analytics to detect and predict threats etc.