“We use data science to automatically classify the crash images, which makes the engineers' everyday work a lot easier. The algorithm recognizes the image content and sorts it into defined categories consisting of individual vehicle modules,” Felix explains. “To do this, we use specially developed AI in the form of a neural network that is trained with existing images from crash tests. This allows us to bypass non-PSW interfaces and implement AI in a risk-aware manner which is in line with compliance requirements.”
Using AI frees up the team's day-to-day work, states Nijaz Dizdarevic, Head of Development Complete Vehicle Crash/Side Protection, “By automatically standardizing our available data, we gain capacity for further development steps in vehicle safety and can work even more efficiently.”
Felix and Nijaz still see a lot of potential in the further use of data science in vehicle development. “There exist numerous other sub-processes that can be automated and I am constantly having discussions with interested team members about potential future steps. For example, a discussion with a co-worker resulted in the idea of using Natural Language Processing to automate part of the requirements management process,” Felix reports. “Precisely because vehicles are becoming more and more connected and more digital, the amount of data available is growing rapidly. Data science can be used both in accident prevention and in damage reduction in the event of an accident. We will be keeping an eye on these technological opportunities at PSW," concludes Nijaz.
