Artificial Intelligence helps design and maintain bridges
ETH Zurich civil engineers are developing new applications of artificial intelligence that can assess the structural safety of railway bridges (Visualisation:Sophia Kuhn and Patankar Yamini/ETH Zurich)
February 24, 2025 – Swiss researchers have developed an AI-based tool to optimize bridge maintenance and extend their service life from the design phase onward. Researchers from ETH Zurich, in collaboration with the Swiss Federal Railways (SBB), have created an AI tool that helps keep bridges in service longer, without excessive material use or increased risk of accidents.
The research consists of two phases. In the first, a model was developed to conduct an initial assessment of structural safety. In the second, an AI assistant was introduced to actively support engineers in designing cost-efficient, safe, and sustainable bridges.
Traditional bridge design relies on conventional calculation software to determine structural safety, usability, costs, and other factors. This process is often iterative, requiring repeated adjustments until project objectives are met a time-consuming approach that leaves significant potential untapped. The Swiss researchers aimed to reverse this process, which was previously impossible with conventional software. However, with ETH Zurich’s newly developed AI assistant, this is now achievable. The tool not only accelerates the design process by evaluating multiple configurations in near real-time but also ensures that proposed designs proactively meet predefined constraints and objectives.
This study builds on research published last year in the journal Automation in Construction under the title ‘Design Space Exploration and Explanation via Conditional Variational Autoencoders in Meta-Model-Based Conceptual Design of Pedestrian Bridges.’
