Integrating AI into Engineering Simulation Workflows

Join us to explore the practical integration, benefits, and limitations of applying AI models within traditional computer-aided engineering (CAE) workflows. Following Siemens’ 2025 acquisition of Altair, Simcenter PhysicsAI and Altair AI Studio now sit within the same unified portfolio — AI Studio itself is now one of the components of Siemens’ new Intelligence Center X platform. This session compares the two underlying approaches (geometric deep learning vs. tabular machine learning), shows how each enables rapid predictions across STAR-CCM+ and HyperMesh, and highlights how PhysicsAI’s outputs can feed directly into Intelligence Center X. We’ll also cover the ROI of implementing AI in simulation environments, along with the data requirements, scaling challenges, and common issues that can shift project scope.

What we will cover:

• Introduction to AI Simulation Tools — core differences between PhysicsAI (geometric deep learning), AI Studio (tabular machine learning, now part of Intelligence Center X), and other options now unified under Siemens.
• Software Integration — a high-level guide to how these tools connect within the combined Siemens ecosystem, including how PhysicsAI-generated data and models can be leveraged in Intelligence Center X.
• Data Generation & Preparation — strategies for extracting functional training datasets using Design of Experiments (DoE).
• Strengths & Capabilities — leveraging AI models for accelerated early-stage design exploration, field result generation, and KPI tracking.
• Core Limitations — risks of extrapolation, model breakdown during abrupt physical regime changes, and the data scaling required for multi-parameter models (curse of dimensionality).
• Evaluating ROI & Practical Applications — identifying which problems justify training an AI model versus relying on traditional full-physics solvers.