Simulation essential to assess Intersection AEB

Virtual simulation essential to assess Intersection Automated Emergency Braking (I-AEB)

Intersection Automated Emergency Braking (I-AEB) is coming to the market (under different names and with different functionality, for example by Audi, Lexus, Mercedes Benz, Volvo) and assessment is on the Euro NCAP roadmap.

In the recent study led by my colleague Ulrich Sander, we investigated whether clustering methods can be used to identify a limited number of test scenarios sufficiently representative of the accident dataset to evaluate I-AEB.

In our study, despite thorough analysis using various cluster methods and variable sets, it was impossible to reduce the diversity of intersection accidents into a set of test scenarios without compromising the ability to predict real-life performance of I-AEB. It was observed, that small changes in the definition of a scenario resulted in a different avoidance outcome, hindering any attempt to group the diverse data. Therefore, extensive virtual simulations are needed to evaluate I-AEB. Nevertheless, physical testing is substantial and necessary: it should be defined to give basic information about I-AEB performance and to verify and validate the virtual simulations.

You can read more about the potential of clustering methods to define intersection test scenarios and to assess real-life performance of I-AEB in our article in Accident Analysis & Prevention Vol. 113, pp.1-11. A preprint version is available.