When a nudge is active and measurements are being made, then comes the time to compare driver behaviour with and without the nudge. It may seem rather straightforward, but it is in fact not. Measuring difference in speed is one thing, measuring whether the speed was appropriate for the situation is a completely different story.

Moritz Berghaus at ISAC, RWTH Aachen will be doing a lot of the data analysis to see if the light nudge in has worked out or not. He will work out a number of parameters, including average velocities, speed distributions and lateral acceleration. But to make some kind sense of the data, it has to be translated into a single safety parameter. A parameter measuring the likelihood of a crash.

At this moment, Moritz and his colleagues are investigating which factors could be included in such a parameter. Except accelerations and configurations of the curve, many other external factors have to be taken into account as well.

Different weather conditions provide very different driving environments and can change numerous factors such as line-of-sight and modified friction for the wheels. Night and day time are not equal, neither are wet and dry roads. But even if the most important parameters are identified, the problem is to connect them to values of how dangerous they are, and which type of behaviour they require.

“It would be great to have lots of data on what happens before a crash,” Moritz Berghaus proclaims. The problem is that accidents are rare, and MeBeSafe is striving to remove them. It would be a bittersweet paradox if we had to rely on crashes actually  happening. “So”, Berghaus says, ”it will be very interesting to see where we will end up with this in the end.”