Studying pedestrian behaviour
Safety engineers study scenarios involving vehicular collision with pedestrians. However, people do not necessarily behave like test dummies. Facing an emergency, pedestrians may suddenly and involuntarily run or jump; this behaviour may put pedestrians in even greater danger, and makes it more difficult to design safety measures. To overcome these problems, an EU-funded study, 'Pedestrian pre-crash reactions and their effects on crash outcomes' (PEDPCREACT), aims to find out how pedestrians react in simulated crash situations and what effect this may have on crash outcomes. A mixed group of youthful and elderly volunteers faced a simulated pedestrian crossing constructed from projections of oncoming two-way traffic. Motion-capture markers were placed all over the volunteers, who faced various startling crossing incidents. The resulting movement and postural data were analysed by computer and used to further simulate car–pedestrian impacts. This enabled pre-crash conditions to be manipulated virtually, showing their effect on the result. The PEDPCREACT study found three types of reaction, which were the same across both age groups, though the proportions varied. Sixty percent of 18–30 year olds accelerated, while the elderly (60–75 years old) froze about as often as they ran (38 % and 40 % of the time). The other response was retreat. Young people also demonstrated reaction times 1.5 times faster than elderly people. For these two actions, the study found that posture varied considerably, except when people froze, in which case they consistently raised their arms to protect their face and head. Pre-impact posture was found to be the most important injury-predicting parameter. Out of 51 simulations at 40 km/h, 10 caused severe head injuries; in 5 simulations, the head injury estimate was higher than the walking posture one, and the rest were lower. This suggests that the standard walking posture used in safety evaluations may not cause the most severe injuries. Thus, setting a testing threshold of head injury indices below that resulting from standard walking posture should eliminate most of the simulated fatalities. This information can be used for more accurate modelling and, ultimately, safer designs.
