| Research Projects | Promoted | AI-Aided Design: Generative AI as a Co-Creation Tool
bidt background

AI-Aided Design: Generative AI as a Co-Creation Tool

The project investigates the development and use of generative AI to support designers in the design of 3D objects that may be realised by 3D printing.

Project description

Generative models represent a development in AI that realises machine-based creativity. Large Language Models (LLM) such as ChatGPT produce high-quality text. Other generative models such as Stable Diffusion or Dall-E create images and videos with unprecedented expression and realism. To this end, a user can specify their goals through textual descriptions, and the system creates a corresponding visual representation. An extension of this concept is to create 3D objects using appropriate generative models, that is, to specify objects verbally and create 3D objects that may be realised by 3D printing, for example.

However, we require effective methods for humans to interact with such a creative machine to steer the creation process toward technically feasible solutions. The long-term vision is to extend, adapt, and develop generative models and machine-based creativity to provide human designers with new tools to facilitate engineering design processes.

The aim of this research project is to implement software methods for the AI-supported design of 3D objects. The objects are designed through a co-creative process. The generative software produces a description of the object in a formal language that can be used to simulate and generate 3D objects.

Project team

Prof. Dr.-Ing. Klaus Diepold

Head of the Chair of Data Processing, Technical University of Munich

David Fresacher M.Sc.

David Fresacher M.Sc.

Research Assistant, Technical University of Munich