Data Science

Predicted performance of airbag designs under load

When is the airbag triggered in a crash, what size and shape does it take in the course of the crash, and how is the pressure in the airbag controlled by means of the vent holes? There are many features which are crucial to the best possible protection of passengers in an emergency, and this detail is a time-consuming part of the vehicle development process. Every parameter change in the airbag system has had to undergo modelling and then simulation hitherto but now simulation results in case of parameter changes can be predicted in real time based on data thanks to a technique developed by Felix Marschall when he did his master’s thesis.

“The method I developed predicts the performance of airbag designs under load based on crash simulations,” explained Felix. “In addition to forecasting the simulation results, we can also use it to make an initial estimate of the effects of parameter changes and search entire parameter spaces for an optimum combination of parameters.” With the locally hosted AI, PSW has noticed a significant reduction in the simulation input and iterative steps during the system development process.

In his role as a student employee and then as a master’s degree candidate, Kevin assisted the Vehicle Safety team in the development of data science methods. “The staff took a great deal of interest in my work and they were always supportive in terms of providing data or discussing further steps in the process,” added Felix. “Data science helps us both in the data-based development of active measures like the targeted deployment of airbags and in passive safety measures, and this makes us even more efficient,” said Nijaz Dizdarevic, Head of Full-Width Frontal Crash & Side Impact Protection Development and supervisor of Felix's thesis.

Previous Article

Soziales Engagement

Next Article

New Work