Distraction beyond the driver: predicting the effects of in-vehicle interaction on surrounding traffic

Recent studies of driver distraction have reported a number of detrimental effects of in-vehicle interaction on driver performance. This paper examines and predicts the potential effects of such interaction on other vehicles around the driver's vehicle. Specifically, the paper describes how computational cognitive models can be used to predict the complex interactions among several vehicles driving in a line when one or more of the vehicles' drivers are performing a secondary task (phone dialing). The results of simulating two distinct car-following scenarios illustrate that in-vehicle interaction by one driver can have significant downstream effects on other drivers, especially with respect to speed deviations relative to a lead vehicle. This work generalizes recent work developing computational evaluation tools for user interfaces in complex domains, and further serves as an example of how user interaction in some domains can have broader effects on the community at large.