SmartUQ Detroit Free Workshop

Modern Design of Experiments, Machine Learning, and Calibration for Simulation and Digital Twins

July 9, 2024 - 10:00 am EDT to 2:00 pm EDT
Suburban Collection Showplace, 46100 Grand River Ave., Novi, MI 48374
Room: Gold/Bronze/Silver Ballroom
Presented by Gavin Jones, Principal Application Engineer

Register

As part of a growing shift toward digital engineering, simulation and digital twins are widely used by many industries including aerospace, defense, automotive, industrial machinery, and semiconductor. However, the effective use of simulation still presents a number of challenges including long run times, uncertainties such as to model form and parameters, and difficulty calibrating to real world data. Join us for this 4-hour SmartUQ Workshop which will introduce tools and techniques for addressing simulation and digital twin challenges including:

• Efficient modern design of experiments (DOEs) for optimally selecting which simulations to perform.
• Training machine learning models (aka predictive models, surrogate models) of simulation to accelerate any workflows or analyses involving simulation.
• Maximizing the knowledge gained from simulation and understanding uncertainties by using machine learning models for sensitivity analysis, uncertainty propagation, and optimization under uncertainty.
• Statistical calibration for addressing parameter and model form uncertainty in calibrating simulations to physical data.

Points will be illustrated through demonstration in SmartUQ and examples from SmartUQ customers in the automotive, aerospace, and semiconductor industries. Free coffee and lunch will be provided, after which the Workshop will conclude with a more in depth training and demonstration of SmartUQ. The workshop is intended to be interactive, attendee questions and participation are encouraged.