AI in Physics-Based Product Performance Simulation: From Surrogates to Solvers

• Slow Workflows: Our traditional “stage gate” simulation workflows are too slow for agile product development, creating significant delays that keep us from innovating quickly.
• ROM/DOE Limitations: We’ve found that reduced order models (ROMs) and DOE techniques are faster than detailed 3D CAD-based simulations but often sacrifice accuracy or require manual setup.
• Expertise Bottleneck: Availability of simulation expertise within product design teams is often a design process limitation, and our efforts at “simulation democratization” have seen only limited adoption.
• Data Reuse Gap: Despite our widespread adoption of PDM/PLM, we as engineers struggle to reuse valuable legacy simulation data effectively, as it is still not part of our “core” data strategy.
• Integration Uncertainty: We lack clarity on how to integrate AI into our long-established simulation workflows, leaving us uncertain about where to start and how to validate the results vs. empirical data.
• Trust and Validation: As engineers and managers, we are concerned about whether AI-driven results can be trusted in our safety-critical industries (aerospace, automotive, energy), which is a major barrier to adoption.

• Understand the emerging AI techniques in simulation: surrogates, PINNs, neural operators, and generative models.
• Learn how startups (BeyondMath, Emmi, Navier, PhysicsX, Neural Concept) and cloud-based simulation solutions (Luminary, Simscale, Rescale) are pushing simulation beyond traditional boundaries.
• Understand how established solution providers (Ansys, Siemens, Dassault, Cadence, Keysight, etc.) are integrating AI into their platforms.
• Gain insight into NVIDIA’s enabling role: PhysicsNeMo (formerly Modulus), Omniverse, and HPC stack.
• Discover how AI, through surrogate models and AI-driven design space exploration, enables generative design approaches that were previously impractical due to computational limits.
• Identify the advantages and limitations of AI vs. traditional ROM approaches.
• Learn about real-world applications across aerospace, automotive, energy, and robotics.
• Recognize the role of enterprise architecture and data management frameworks in enabling this transformation.

This webinar is designed for anyone involved in systems engineering, product development, and simulation strategy and best practices implementation, including:

• Simulation & analysis engineers (CFD, FEA, multiphysics).
• R&D managers who are seeking faster design cycles and innovation.
• Digital twin and AI/ML specialists exploring hybrid models.
• IT/HPC decision-makers evaluating GPU or cloud deployment strategies.
• Executives assessing the impact of AI in simulation on competitiveness and ROI.