Structural Design Codes with AI: From Safety Margins to Smart Design

This presentation explores how Artificial Intelligence can transform traditional structural design codes, such as the Eurocodes, by enhancing accuracy, reducing conservative safety factors, and enabling optimized designs. AI, especially physics-informed models, offers the ability to predict and distinct complex failure mechanisms and their combinations that elude simplified mechanical models—particularly in thin-walled steel structures and steel-concrete composite systems where multiple or interconnected failure modes may occur. These insights are vital as the industry shifts toward modern methods of construction, such as modular systems, where adaptability and predictive intelligence are key. AI agents and assistants can act as design co-pilots, streamlining code interpretation and adaptation for projectspecific optimizations. Computer vision and pattern recognition are employed to assess the structural components— paving the way for a circular economy in construction. Moreover, AI enhances predictive modelling accuracy and provides real-time assistance in delivering higher quality, code-compliant structures that reflect the real-world behavior of materials and connections more faithfully.