Description
Explore how you can use artificial intelligence (AI) to extend the reach and accessibility of catchment-specific machine learning (ML) in InfoWorks ICM software without sacrificing hydraulic fidelity. This session will introduce the concept of training surrogate ML models on local InfoWorks ICM simulations to predict flows at key locations across a catchment. You’ll learn how Autodesk is generating and deploying these surrogates to quickly forecast system performance without computationally expensive simulations. The approach offers a new way to support rapid scenario testing, stakeholder engagement, and real-time operational insight without hurting a hydraulic equation. The session will present a practical framework for integrating AI in a way that empowers more users and accelerates insights.
Key Learnings
- Learn how surrogate models are trained on InfoWorks ICM results to replicate catchment behavior.
- Learn how Autodesk is deploying machine-learning surrogates to predict flow performance at key locations in real time.
- Learn the trade-offs and benefits of integrating AI to replacing core hydraulic modeling principles.
- Learn about best practices for training surrogate models—including rainfall input—to ensure reliable flow predictions.