Enhancing Stealth Gameplay Through Procedural Generation: An Algorithmic Approach to Dynamic Guard Paths and Placement in Infiltration Games

Infiltration or stealth-based games are a signifi-cant genre in the gaming industry, centered around undetected navigation through levels, avoiding guards, cameras, and other security mechanisms. Level design for such games is a complex challenge, raising questions about guard placement, patrol routes, player viability, level difficulty, and overall player experience. Despite its importance, very few scientific works addressed the challenge in the past. It is why this paper introduces an innovative procedural generation technique for automatically de-signing maps, positioning guards, determining their patrol paths, and evaluating level difficulty, with broad applicability across various game types. Implemented in Unity 3D, our method was validated by generating over five hundred maps, demonstrating its effectiveness and the high quality of the generated levels.