On the effective co-creation of CAD models by leveraging generative artificial intelligence

Collaboration between humans and generative artificial intelligence (GenAI) can result in effective problem-solving methods for smart manufacturing. From this point of view, this study presents how to solve computer-aided design (CAD) problems using GenAI-human collaboration, where a GenAI tool generates structured code (syntax) for CAD while a designer articulates the design intent (semantics). The focus is to see how the information type, model complexity, and prompt structure collectively affect the collaboration. A set of case studies demonstrate that a modular modeling approach with low-level information improves prompt efficiency and modeling accuracy, especially in complex scenarios. This strategy also enables novice and expert users to collaborate with GenAI in solving challenging real-world CAD tasks effectively. The findings support the development of advanced human-AI co-creation systems and encourage future research in areas such as reverse engineering, multi-part assemblies, and collaborative CAD workflows coupled with complex design constraints.