Analysis of artificial intelligence applications for automated testing of video games

Abstract

Game testing is a software testing process for quality control in video games. Game environments, sometimes called levels or maps, are complex and interactive systems. These environments can include level geometry, interactive entities, player and non-player controllable characters etc. Depending on the number and complexity of levels, testing them by hand may take a considerable effort. This is especially true for video games with procedurally generated levels that are automatically created using a specifically designed algorithm. A single change in a procedural generation algorithm can alter all of the video game levels, and they will have to be retested to ensure they are still completable or meet any other requirements of the game. This task may be suitable for automation, in particular using Artificial Intelligence (AI). The goal of this paper is to explore the most promising and up-to-date research on AI applications for video game testing to serve as a reference for anyone starting in the field.