Publication

This report provides an overview and an in-depth understanding of the characteristics of road traffic crashes involving vehicles (here focussing on passenger cars) and vulnerable road users (VRUs, i.e. pedestrians, cyclists, riders of mopeds, e-bikes or scooters), primarily in European countries. Several crash databases including international, national and in-depth crash information have been analysed. Among them, the CARE database (Europe), the German, Swedish and Hungarian national road traffic statistics, as well as the in-depth databases IGLAD (Europe), GIDAS (Germany), in-depth data from Pest county (Hungary) and the Volvo Cars Cyclist Accident Database (Sweden). Early investigations have shown that the crashes between passenger cars and pedestrians or cyclists are most frequent in Europe, hence these crashes were investigated in greater detail. Further, this report describes briefly the road safety situation of scooter riders in collisions with passenger cars and crashes between trucks and VRUs. As the structure of the databases was quite different, not all results for different countries could be compared directly. Nonetheless, trends could be identified from the analysis. Considering cyclists the highest numbers of fatalities per inhabitants can be observed in countries where cycling is very common and the bicycle is used as a daily transportation means like in The Netherlands and in Denmark. Similar to the observation for pedestrians in former projects, elderly people have the highest risk to get fatally injured as cyclist riders in most countries due to their high vulnerability. Overall, datasets confirmed that older cyclists suffered more often from higher injury severities compared to younger ones, male cyclists were injured more often than females, higher injury severities (in particular fatal crashes) happened more often on rural roads and that crashes occurred most often in fine weather and daylight conditions. The in-depth understanding of the crashes included the identification of the most relevant road traffic ‘accident scenarios’ and levels of injury severity sustained as well as the transport modes that represent a higher risk for VRUs. Within PROSPECT, an ‘Accident Scenario’ is described by the type of road users involved in the accident, their movements (e.g., the motion of the cyclist or pedestrian relative to the vehicle) expressed as ‘accident types’ and further relevant contextual factors like the course of the road, light conditions, weather conditions and view obstruction. The wording ‘Target Scenario’ or ‘Use Case’ is often used to describe ‘target groups’. Within PROSPECT, ‘Target Scenarios’ are equivalent to ‘Use Cases’. They are derived from accident scenarios by adding more detailed information about the road layout, right-of-way, as well as manoeuvre intention of the driver. One accident type can be split into several Use Cases. egarding car-to-cyclist crashes, it was concluded to consider five Accident cenarios: (I) “Car straight on, Cyclist from near-side”, (II) “Car straight on, Cyclist rom far-side”, (III) “Car turns”, (IV) “Car and cyclist in longitudinal traffic” and (V) Others”. Focusing on killed and seriously injured (KSI) cyclists, results for Germany, ungary and Sweden were similar regarding scenarios (I) and (II); around 42%-52% f all casualties were assigned to these scenarios. However, the results varied a lot etween the considered countries for Accident Scenarios III and IV. In particular, ungary seems to have major issues with cyclists in longitudinal traffic compared to ermany and Sweden, which could also be caused by infrastructural differences. ocusing on killed cyclists in car-to-cyclist crashes, it can be seen that in all countries he accident scenario IV (longitudinal traffic) had the greatest relative frequency of all ccident scenarios ranging from 25-64%. This was linked to the higher car impact peeds observed on rural roads. se Cases have been derived from these Accident Scenarios for car-to-cyclist as ell as for car-to-pedestrian crashes. Major parts of this work have been published in separate PROSPECT papers and deliverables. [...]