This research introduces an automated approach to creating automotive domain models using GPT-4o, streamlining a traditionally manual and complex process.

• Converts automotive requirements directly into Ecore metamodels
• Implements an iterative refinement process for model accuracy
• Visualizes results using PlantUML notation for improved accessibility
• Deployed as a web service for practical industry application

Industry Impact: This approach significantly reduces the engineering effort required for automotive system modeling and standardization, enabling faster development cycles and improved collaboration between technical teams.

LLM-based Iterative Approach to Metamodeling in Automotive