Computational Intelligence-Driven Innovation in Vehicle Body Design Automation

This paper presents a Computational Intelligence-Aided Design (CIAD) framework for accelerating early-stage automotive body development through the integration of artificial intelligence, generative 3D modeling, parametric CAD, and automated engineering simulation. The proposed system employs an AI agent built with Ollama and ComfyUI to generate initial 3D vehicle geometries directly from natural language descriptions. These mesh models are then converted into solid parametric models using Autodesk Inventor and evaluated through simulation-driven analysis in Ansys Workbench. Python scripting automates the interaction between AI generation, model refinement, and performance validation, enabling a seamless, closed-loop design process. By bridging the gap between conceptual design and engineering feasibility, CIAD offers a scalable solution for rapid, iterative vehicle development. Future work will focus on broader validation across diverse case studies and comparative performance evaluation to support adoption in industrial practice.