Curve speed warning systems (CSW) utilize information about the road and warn drivers if they are about to enter a curve too fast. Recent research shows that CSW is successful in warning for upcoming curves. However no statistically significant change in driver behaviour due to CSW has been shown. In addition, a common requirement cited by drivers is that the amount of false alarms needs to be reduced.

This paper evaluates how the level of detail in the modelled vehicle dynamics influences the threat assessment in a situation with an oncoming curve. The point mass model that is commonly used by CSW is compared with more detailed models. Maximum velocity the vehicle can have while still following a curve is investigated and compared for the point mass model, the single track model and the double track model. It is shown that as the level of detail in the modelled dynamics increase, the maximum velocity profile is significantly reduced. This implies that in order to make a reliable threat assessment that can reduce the amount of false alarms and even be used as a base for an autonomous intervention, a more complex vehicle model than the point mass model is required.