An Exploration of Collision-based Enemy Morphology Generation

2026-06-01

AI Summary

This study addresses a notable gap in procedural content generation (PCG) research by focusing on the automatic generation of enemy collision structures in video games—a domain largely overlooked in prior work. The authors propose three novel methods for generating enemy body and collision geometries by leveraging player collision data, integrating procedural content generation, collision data analysis, and evolutionary algorithms. Experimental results demonstrate that the proposed approaches either outperform or match evolutionary baselines adapted from robotics literature. These findings substantiate the feasibility, effectiveness, and diversity of generating enemy morphologies grounded in empirical collision information, thereby offering a new direction for AI-driven content creation in game development.

Abstract

Despite a great deal of prior research into Procedural Content Generation (PCG), relatively little prior work has explored generating enemies for video games. In particular, there is almost no work on generating enemy morphologies, the basic body plan or collision information for in-game enemies, despite the existence of related morphology generation work in robotics. In this paper, we explore three different novel approaches to generate enemy morphologies based on player collision information. We found that each approach provides different strengths and weaknesses, but all had equivalent or better performance than an evolutionary baseline adapted from prior robotics morphology work.

Problem

• Procedural Content Generation
• Enemy Morphology
• Collision-based Generation
• Video Game AI
• Morphology Generation

Innovation

• Procedural Content Generation
• Enemy Morphology
• Collision-based Generation
• Video Game AI
• Morphology Design