Generating video game puzzles through planning

Planning is an AI concept that is well-known in the video game industry and has been applied to video games. A large number of these applications have focused on using planning to control the behaviours of agents within video games. However, there is comparatively little research about the application of planning in video games for non-behavioural AI; that is the focus of this work. Many video games consist of puzzles or puzzle elements that players have to solve. Puzzles have a defined search space that can often be considered a planning domain, and planners provide a useful tool for finding a valid series of actions that can efficiently solve such problems. The aim of this work is to explore whether it is possible to create valid and challenge-appropriate puzzles for players in video games through automated planning.