Assessing Red Light Running Behaviors and Their Consequences: Two Complementary Studies Based on the Nantes Area
L Carnis, IFSTTAR FR; R Dik, E Kemel, CETE de l'Ouest, FR
The study consists in both a quantitative assessment of red light offences and accidents, and a qualitative analysis of these accidents. This approach provides an overall look on this kind of offences and the related consequences.
Although France targets to reduce the number of road fatalities under 3000 by 2012, about 4000 lives have still been claimed on French roads in 2010. New niches of traffic safety thus remain to be found. A direction currently explored is the use of automated traffic rules enforcement. The implementation of an automated speed camera programme in 2003 represents one main example of this new public policy. More recently, automation of enforcement has been expanded to other traffic offences. As a matter of fact, authorities deployed in 2009 the first red light cameras.
In this context, it is relevant to assess red light offences and the related accidents in order to evaluate to what extent changing traffic light running behaviors can improve traffic safety. This paper proposes such an analysis based on the Nantes area. Data on red light running cases have been collected over a one year period on 26 sites representative sites. The data includes the hour and the day of the committed offences as well as some information about the traffic. More than 56000 offences were recorded for about 3.6 million crossings.
The first aspects of the analysis consisted capturing the patterns of red light offences, using econometric tools. The study then consisted in collecting information about traffic accidents that occurred around traffic lights in the same area over the 2003-2009 period. This data allowed to match accident patterns with red light running ones. In addition to the time and location of these accidents, a particular attention was paid to the type of road users involved. Lastly, a deeper look was given to 30 complete descriptions of accident that occurred on four routes, selected for their high linear density of traffic lights. Such data offered a qualitative understanding of the factors that can lead to traffic light accidents.
Interestingly, red light running behaviors are not only strongly related to the time of the day, a significant effect was also observed depending on traffic conditions and urban context. This suggests that, beyond individual propensity to abide to traffic rules, environment is a key factor for this behavior. Such factors are also found to be relevant to characterise traffic light accidents.
Coupling a quantitative assessment of red light offences and accidents, with a qualitative analysis of these accidents, the paper provides an overall look on this kind of offences and their consequences. Among other it shows which types of accidents can be avoided thanks to a better knowledge of red light running behaviors.
Association for European Transport