The safety effects of Daytime Running Lights

A perspective on Daytime Running Lights (DRL) in the EU: the statistical re-analysis and a meta-analysis of 24 independent DRL-evaluations as well as an investigation of possible policies on a DRL-regulation in the EU
Auteur(s)
Koornstra, Drs. M.J.; Bijleveld, Drs. F; Hagenzieker, Drs. M.P
Jaar
History The original reason for the use of Daytime Running Lights (DRL) was not the improvement of vehicle visibility. The use of DRL seems to have originated in 1961 as a campaign to operate motor vehicles with headlights on in daytime as a signal of the intention to comply with a Texas governor's request to drive safely. Also at that time, to quote a remark made by investigators of a DRL effect in the early sixties: `It seems that no one can conceive of an automobile or a Greyhound Bus being invisible on a bright clear day'. This view is applicable to most road users even today. Research methods and conclusions In this study the role of perception in accidents and the effects of the introduction of DRL have been reviewed together with all 24 existing evaluations of DRL. Additional statistical analysis and new techniques have been employed to produce the best estimates possible of the full effects of the introduction of DRL in the EU in terms of the saving of lives and reducing the costs of the road transport system. DRL as a road safety measure is often difficult to understand for the road user because he or she `knows' that with sufficient attention every road user can be seen in daylight. Nevertheless, the research reviewed shows that visual perception in daytime traffic is far from perfect and it is worse in conditions of low ambient illumination. In a striking example 8% of cars in an open field in broad daylight were not visible from relevant distances without the use of DRL. On shady roads or those with backgrounds which mask objects in the foreground the visibility and contrast of cars in popular colours is greatly reduced. It is known from in-depth accident studies that failing to see another road user in time (or at all) is a contributing factor in 50% of all daytime accidents and for daytime intersection accidents this increases to as much as 80%. The psychological research reviewed shows that DRL does not only improve the visibility of motor vehicles in daytime, but also influences the timely peripheral perception of vehicles making conflicting movements. Moreover, cars with DRL are better identified as cars and their distances are estimated more safely compared to cars without DRL. All this contributes to the expectation that DRL has positive safety effects, especially in conditions of low ambient illumination. However, until recently, even road safety scientists debated the validity of DRL effects in other conditions than in Nordic winter daylight. The scientific evidence for the safety effects of DRL in latitudes to the south of the northern Scandinavian countries has only become available recently (Denmark, Hungary, Canada). Older DRL-evaluations for southern regions mainly concerned DRL for company fleets in the USA, but results, though positive, were not statistically significant. New meta-analysis of the earlier and more recently available DRL-studies, taken together, have now shown that DRL-effects on the same latitudes as those applicable to Europe are statistically significant. This study investigated for the first time the differences between national and company fleet DRL-effects as well as the DRL-effects on accidents and on casualties. Both are found to be statistically significant. In this study all existing (24) independent DRL-evaluations have been reviewed and/or re-analysed in order to obtain unbiased, and comparably defined, intrinsic DRL-safety-effects while estimating statistical uncertainties in an optimal way. Intrinsic DRL-safety-effects are defined as the effects of a change from 0% to 100% use of DRL by motor vehicles. The observed effects of DRL will differ, therefore, from the intrinsic effect when DRL usage is not zero at the start and/or not 100% at the end of observations. The intrinsic DRL-effects calculated in this study cover 9 countries and are combined into 12 national intrinsic DRL effects, 5 on multiple (multi-vehicle) daytime accidents and 7 on casualties in multiple daytime accidents. The result of this analysis is the establishment of statistically significant curvilinear relationships between latitude and national DRL-effects with respect to both accidents and casualties. From the difference between these two relationships an estimate has been made for the relationship between latitude and DRL-effects on fatalities in multiple daytime accidents. Figure 1 displays these relationships and the 12 national intrinsic DRL effects
Rapportnummer
R-97-36
Pagina's
175
Gepubliceerd door
SWOV, Leidschendam

SWOV-publicatie

Dit is een publicatie van SWOV, of waar SWOV een bijdrage aan heeft geleverd.