To get a sense of an external distraction that might be compared to the response to UAVs in traffic, it seemed appropriate to look at studies done on digital or video billboards. The movement on these screens has been shown to draw drivers’ eyes more often and for longer periods than traditional static billboards. One study showed this comparison, concluding that drivers passing video advertisements:
In a study on the Gardiner Expressway in Toronto, infrared eye-tracking devices also measured the tendency of drivers to glance at both passive (non-movement) and active signs and billboards. They found that the active signs – those including video, scrolling text and partially active roller bar signs – received an average of 1.31 glances per subject per sign, as opposed to 0.64 glances per subject per sign for the passive signs (Beijer, Smiley & Eizenman, 2004).
A study on in-vehicle glance duration and its impact on crashes showed that glances away from forward for more than 2 seconds showed increased risk of crash (Horrey, 2007). With that in mind, Divekar et al.(2012) conducted research with a simulator that looked at distraction from a digital billboard, coupled with another task requiring attention. Eye movements of the participants and measures of the vehicle, such as lane position and speed were measured. Both experienced and inexperienced drivers were tested; the average maximum glance away from the forward roadway was 3.12 seconds for the inexperienced group vs. 2.96 seconds for the experienced group. The study also showed that they generally looked away with one long glance rather than a series of short glances. When coupled with a hazard anticipation task, the distraction was more evident. A passive hazard test measured whether or not the driver scanned a line of parked cars. With other external distractions (digital billboard), the hazard anticipation of the novice group dropped from 37.5% to 8.3%. The experienced group dropped from 38.4% to 15%. An active hazard scenario had a pedestrian stepping from behind a tree, and participants were scored on whether or not they detected the pedestrian. Without the external distraction of the digital billboard, hazard anticipation for novice and experienced drivers was 87.5% and 95% respectively. With the introduction of the external task, these percentages dropped to 37.5% and 52.6%. Another simulator study by Chan et al. (2008) with novice and experienced drivers performing secondary external tasks reported even longer average glances away from forward - 3.75 seconds for novice; 3.42 seconds for experienced. While these averages are significantly longer than the accepted threshold, another study showed that even glances ≥ 0.75 seconds from forward could significantly increase the likelihood of crashes when there are cars within 36 m (118 ft.) (Beijer, Smiley & Eizenman, 2004).
Young et al., in a 2009 simulator test, found that drivers were more likely to recall the advertising billboard rather than the traffic signs, perhaps indicating another layer of distraction.
Studies have made a definite correlation between external-to-vehicle distractions and increased potential for crashes, a point that Wallace (2003) says is supported by by accident databases. He warns, though that the distraction factor may be significantly under-reported.