New databases

At present, there are some 50,000 fatally injured road users each year throughout the 25 EU Member States. A core element of the EC road safety strategy includes a reduction of fatalities by 50% by the year 2010. Central to this strategy is the requirement for good quality in-depth accident data. Such data are seen as a fundamental pre-requisite for the formulation and monitoring of road safety policy in the EU. Data are needed to assess the performance of road and vehicle safety stakeholders and to support the development of further actions. A recent analysis conducted by the European Transport Safety Council identified that no single accident database could meet all of the needs and that there were major gaps including better knowledge about in-depth accident causation. Specific policy questions at EU level involve the role of road infrastructure in accident causation, the monitoring of progress towards the 2010 targets, and the improvement of vehicle design and performance in accident and injury causation. In view of these considerations, one of the aims of the SafetyNet project is to provide both an in-depth fatal accident database and a (more detailed) in-depth accident causation database.

In-depth fatal accident database

Task 5.1 of the SafetyNet Work Package 5 is developing a broad ranging, intermediate level, fatal accident database. The dataset is systematically selected according to a defined sampling plan and the data are representative of the countries in which the data are collected. The data collection areas for the accidents will be from the countries with the largest fatality populations in Europe (Italy, France and Germany) as well as from northern European (Sweden, Finland) and from middle European (UK, Netherlands) countries. Independent groups with no interest in commercial aspects of the study outcomes conduct all data gathering and accident investigation activities.

The data recorded describe the environment (including road infrastructure, e.g. crash barriers, road signs), vehicle, and driver factors to provide a description of the whole crash. The database will strive for a level of detail similar to FARS database and Stats19. Around 100 – 150 variables for each case (accident/vehicle/ occupant/other records) include about 500 pieces of information per case.

In-depth accident causation database

The purpose of the independent in-depth accident causation data is to put together a crash investigation process that identifies the main risk factors leading to a crash. The independent in-depth accident causation database will have applications in the areas of new technology development, active safety systems, infrastructure, and road safety. It will be a new accident investigation activity and the in-depth data will have significant applications for policy making and road safety practitioners, particularly for those working with infrastructure safety.

The data collection areas for the accidents are from the countries with the largest fatality populations in Europe (Italy, France and Germany) as well as from northern European (Sweden, Finland) and middle European (UK, Netherlands) countries. Independent groups with no interest in commercial aspects of the study outcomes conduct all data gathering and accident investigation activities. The following organizations participate with investigative teams:

• VSRC (Vehicle Safety Research Centre)
• TNO (Netherlands Organisation for Applied Scientific Research)
• Chalmers (Chalmers University of Technology)
• MUH (Medical University of Hannover)
• VALT (Finnish Motor Insurers' Centre)
• DITS (Department ‘Idraulica Transporti Strade’ University of Rome).

Specialist teams investigate the causes of accidents in six countries to give a survey of those Member States. Although it will prove hard to collect data from a random selection of cases, there is an aim for randomness within the framework given. This means that the selection of cases is kept random to the best of the teams abilities.

There has been a general consensus that critical events should be used as the strategy for this project. A categorization of variables has been made and the variables are divided into six different groups:

• general variables
• critical events
• contributing factors involving the road user
• contributing factors involving the vehicle
• contributing factors involving the infrastructure
• contributing factors involving the organisation