Generative bridge design
Table of contents
Project data
|
Titel | Title |
Short description
The bridge design process is time-consuming, costly, and challenging to standardize, as each structure must be adapted to specific conditions and requirements. Despite engineers’ reliance on technical knowledge, guidelines, and experience, variations in design and inconsistent quality persist. Generative AI offers a completely new design approach that facilitates both high levels of automation and the integration of physical mechanisms, while also incorporating global knowledge and experience to benefit from the best designs worldwide. This research aims to advance generative AI-based bridge design by integrating data, engineering knowledge, and experience to create material-efficient and structurally effective solutions. We will develop an interactive, multimodal generative AI framework that leverages deep learning (DL), reinforcement learning (RL), and physics-informed loss functions to integrate human expertise, enabling the computer-aided generation of high-quality bridge designs.