Agentic AI for 2D Legacy to 3D CAD Model Conversion

Computer-Aided Design (CAD) models are essential for modern manufacturing, serving as the basis for generating instructions for CNC machines, 3D printers, and other automated systems. However, many legacy designs are preserved only as 2D engineering drawings with annotations rather than complete 3D models. Reconstructing accurate 3D CAD models from these legacy drawings typically requires skilled interpretation using parametric modeling software, making the process labor-intensive, time-consuming, and susceptible to errors that can result in incompatible parts or manufacturing defects. As advanced manufacturing increasingly relies on digital 3D models, there is a growing need for efficient and reliable techniques to reconstruct 3D CAD models that accurately preserve the fit, form, and function of legacy designs.

 

Researchers at Arizona State University have developed an AI-driven solution that automates the conversion of 2D legacy CAD files into accurate 3D models using advanced language and vision models. This technology leverages Agentic AI by integrating Large Language Models (LLMs) and Vision Language Models (VLMs) to autonomously interpret and performs 3D CAD model reconstruction with multimodal auxiliary information. It eliminates the need for manual, time-consuming processes traditionally performed by engineering experts, providing a streamlined, efficient, and precise workflow for CAD model reconstruction. This technology utilizes a modular architecture and procedural approach making it generalizable and easy to adapt to domain-specific scenarios such as aerospace, automotive, mechanical engineering, and industrial design.

 

This method to construct 3D models from 2D models using artificial intelligence provides a streamlined and precise workflow that can be used in a multitude of industries.