Shouldn ’ T Cars React as Drivers Expect ?

The objective of this project is to develop and test a multi-method empirical approach for predicting drivers’ assessments of the level of acceptability of a warning issued in response to accidents, near-accidents, and other incidents. The role of humans (drivers) in the pre-crash phase means that systems that protect occupants and pedestrians must be seen as distributed, cognitive systems. Driver acceptance therefore has to be an important design goal. One obstacle to acceptance is the human dislike for false alarms. An approach to overcoming driver dislike for false alarms is to focus on driver expectations and to design systems to issue alarms when and only when the driver is likely to accept them. In this paper we discuss one such approach.