DAAAD Bridges: Domain-Aware-AI Augmented Design of Bridge Structures

DAAAD Bridges is a project that aims to develop an AI-assisted parametric design framework to aid in making structural design decisions.

This framework will act as an active, intelligent partner or "Co-Pilot" in an interactive design process. The conceptual design stage in the Architecture, Engineering, and Construction (AEC) industry often requires more performance evaluations, primarily relying on the designer's expertise. Integrating AI technology promises to create more efficient and sustainable structures.

Project Description: The DAAAD Bridges tool is an AI assistant that uses modern neural network architectures to revolutionize structural design processes. One of its most significant features is the "inverse design" capability, which traditional simulation tools lack. With this function, designers can set project goals, enabling the AI model to generate customized, high-performance designs in real time. This capability facilitates the creation of performance-based designs and speeds up their evaluation, exploration, and optimization.

The tool's functionality and benefits are demonstrated through various case studies, mainly in bridge design. Bridge projects are inherently complex due to the vast number of interconnected design parameters, multiple conflicting objectives, and the involvement of numerous stakeholders. This complexity often results in a time-consuming, iterative design process, where the AI assistant can significantly reduce the need for costly alterations in later project phases. Our case studies include a real-world pedestrian bridge design project in St. Gallen and the structural safety evaluation of aging Swiss concrete railway bridges. Collaborations with industry and academia further demonstrate the tool's broad applicability, as seen in projects such as the generative design of timber and steel grid-shell structures and the design of timber connections for high-rise buildings.

The future looks even brighter as we explore integrating the AI assistant into virtual reality (VR) environments. This integration will enhance the collaborative design experience between human designers and AI technology, which we believe will be a game-changer in the industry. We are proud of what we have achieved and look forward to developing this research.