PCGLab3D: laboratório de geração procedural de conteúdo em Godot

Abstract

Introduction: Procedural Content Generation is a well-established approach in game

development that uses algorithms to create environments, items, and other game elements,

reducing production costs for developers while increasing gameplay variety for players. Objectives: This work presents the migration of the PCGLab tool to a widely used game engine, aiming to provide an environment suitable for both academic research and industry applications. In addition, it quantitatively and qualitatively analyzes two strategies for placing elements in roguelike rooms. Materials and Methods: The original environment, along with its generation methods based on influence maps, was adapted to Godot Engine 4. The numerical analyses and designer feedback charts were recreated,

and new ones were added. A new data export feature was also implemented, allowing numerical analyses to be performed outside the tool. Results: This work introduces PCGLab3D, a new laboratory for Procedural Content Generation research. It supports the generation, analysis, and testing of procedural dungeons in a discretized three- dimensional environment, where metrics are collected using influence maps and player paths that simulate the player's experience. This provides both industry and academia with a new tool for creating and comparing different content generation methods.