Optimising Urban Flood Resilience
A novel AI system couples a hydrodynamic model with a custom algorithm to design optimal flood infrastructure.
A team of researchers, including James Mckenna and Christos Iliadis, has published a paper detailing a novel AI-driven tool designed to optimize urban flood resilience. The system addresses the critical challenge of implementing sustainable Blue-Green Infrastructure (BGI)—like rain gardens and permeable pavements—by coupling a high-fidelity, fully dynamic hydrodynamic model with a custom-built evolutionary algorithm. This combination allows the tool to accurately simulate flood water behavior at a property scale, evaluating both hazard and vulnerability, a significant upgrade over simpler models that only predict flood extents.
The core innovation lies in the bespoke evolutionary algorithm, which is specifically engineered to minimize the number of computationally expensive hydrodynamic simulations required. This makes exploring a vast, intractable search space of potential BGI configurations computationally practical. The tool was validated against benchmark algorithms, proving its efficiency. For city planners and engineers, this translates to an automated system that can rapidly generate a portfolio of optimal, cost-effective BGI designs, moving beyond traditional trial-and-error methods to enable data-driven, informed investment decisions for climate adaptation.
- Couples a high-fidelity hydrodynamic model with a custom evolutionary algorithm for accurate flood simulation.
- The bespoke algorithm minimizes simulations, making optimization computationally practical for complex urban environments.
- Automates the design of Blue-Green Infrastructure, providing planners with a set of optimal, cost-effective solutions.
Why It Matters
Provides cities with a data-driven AI tool to design cost-effective flood defenses, crucial for climate adaptation planning.