Understanding and improving the street crossing decisions of elderly pedestrians
The elderly people form a particularly overrepresented group in pedestrian accidents in many European countries. In France, in 2012 persons aged over 75 years accounted for 41.1% of pedestrian fatalities, while they make up only about 9% of the country’s population (ONISR, 2012). This age group also accounts for more than 20% of those with injuries requiring hospitalisation, i.e. 805 persons in France in 2012. While accidents often have fatal consequences for an elderly person, they can also have severe nonfatal consequences, sometimes leading to a total loss of autonomy.
In spite of the importance of the topic for road safety, the literature contains little research that specifically investigates the decisions and behaviours of elderly pedestrians. Progress in this area has only been made in the last 20 years.
What difficulties are involved in crossing a road?
To understand why elderly pedestrians are often involved in accidents, research has been conducted in situations where pedestrians receive no assistance (for example where there are no pedestrian traffic signals) and where they can make the decision on their own to cross the road. These studies show that individuals of over 75 years of age often select time intervals that are too short considering their slow walking speed. This has been observed on simulators (Dommes & Cavallo, 2011; Lobjois & Cavallo, 2007, 2009; Oxley et al., 2005), on videos of real traffic (Holland & Hill, 2010), and in observations of real situations (Oxley et al., 1997).
This work reveals that elderly persons are slower to make decisions and take the first step, walk more slowly, have diminished abilities to accelerate, and, above all, difficulties perceiving the speed of approaching vehicles or assembling the information from two traffic lanes (Dommes et al., 2014). Generally, the increasing danger faced by elderly pedestrians is interpreted as resulting from the decline in certain capacities that occurs with normal ageing rather than the age of the pedestrian as such (Dommes et al., 2013; Holland & Hill, 2010; Lobjois & Cavallo, 2007, 2009; Oxley et al., 1997, 2005). Such decline affects visual acuity, perception (effective visual field), cognitive activities (attentional capacities), and motor activities (walking speed).
How can decision-making be improved?
One way of improving safety is to make elderly persons better able to cope with the infrastructure and tasks by learning or re-learning safe behaviours. The development of training programs that are specifically targeted at elderly pedestrians would be an effective way of improving safety for this group of vulnerable road users.
Studies conducted at IFSTTAR (Dommes et al., 2012; Dommes & Cavallo, 2012) are, to our knowledge, the first to have tackled this question. The results from simulator tests of a number of training programmes are very promising.
They show that elderly persons’ decisions and behaviours can be made significantly safer by a combination of individualised behavioural and educational interventions. The programmes in question present a variety of exercises of increasing difficulty and give the subjects feedback about their errors which improves their motivation and involvement throughout the intervention.
Virtual reality, an effective technique for studying and improving behaviours
Methodological difficulties related to studying road crossing in real situations have restricted the analysis of risk factors and causal factors for accidents. An accident analysis has been carried out (Fontaine & Goulet, 1997), typical scenarios have been identified (Brenac et al., 2003), but some questions remain unanswered. Advances in simulation have increased our knowledge in this area and its applications, and will continue to do so in the future.
IFSTTAR (LEPSIS) is currently the only institute in the world to possess a road crossing simulator in which the pedestrian really moves over a distance of more than 7 metres. The subject is fitted with movement sensors that allow their position in virtual space to be identified and to adapt the visual scenes projected on 10 large screens to the subject’s movement and point of view which change when they walk.
The street crossing simulator developed by IFSTTAR
