Automotive engineers can use emerging ML methodologies to speed up the product engineering process.

The automotive product engineering process involves months or years of iterative design reviews and refinement, with back-and-forth feedback between stakeholders regularly to adjust designs and evaluate the impact of design changes on engineering metrics like the coefficient of drag. Between each design iteration, engineers wait hours or days for simulations to complete, which means they can only execute a handful of design decisions each week.

In this post, we’ll show how automakers can reduce cycle times from hours to seconds by leveraging surrogate machine-learning (ML) models in place of HPC, physics-based simulations and create subtle design variations for non-parametric geometries.

Read the entire blog post here: aws.amazon.com/blogs/hpc/using-machine-learning-to-drive-faster-automotive-design-cycles/