Where surrogates are used / Across industries

Design use

Millisecond answers turn design studies, optimisation and real-time tools into routine work.

• Design exploration: Sweep thousands of variants in the time one CFD run takes, and find the promising region before spending compute.
• Optimisation: Put the surrogate inside an optimiser. A differentiable surrogate also gives gradients with respect to the design.
• Uncertainty: Propagate manufacturing tolerances or flight-condition scatter by Monte Carlo, which is impossible with a solver in the loop.
• Real time: Interactive design tools and digital twins that keep up with a slider, or with sensors on the real asset.

Related

• Limits and trust: A surrogate is only as good as its data: know the domain, quantify error, verify the winner.
• Where it fits: Tests validate simulations, simulations train the surrogate, the surrogate picks what to simulate next.
• Get started: From open frameworks and datasets to a surrogate trained on your own simulations.