Magna is the leading supplier in the automotive industry, with operations in more than 28 countries and over 100,000 employees. They design and manufacture advanced systems found in nearly every major automotive brand. Magna’s vision is to continue advancing mobility for everyone and everything.

Magna relies heavily on a wide array of component models. These components exist in many variants across different suppliers. Engineers must view, examine and compare the models to further develop their products. This task is slow and difficult to complete, especially as the scale of models increases.

Our LLM 3D Model Interpretation and Decomposition tool eliminates the slow and tedious process of analyzing these component models. The system automatically interprets 3D models, decomposes them into meaningful sections, and generates clear descriptions of each part and its structure. Variants are placed side by side in an interactive three-dimensional viewer where key differences and similarities are highlighted in real time.

Engineers can also query for specific parts using a natural language query through plain text, removing the need to filter and search through thousands of models, turning hours of work into mere seconds. Users can find any part based on any criteria such as material, supplier, or even a specific functionality of the part.

Our system streamlines the automotive component comparison process, saving time and money.

Our LLM 3D Model Interpretation and Decomposition tool is developed as a Python-based pipeline with Open3D geometry processing, vector storage through Neo4j, and an AI-powered FastAPI back end. The results appear in a responsive web application with visualization powered through Three.js, for detailed analysis.