Overview
The number of road fatalities in Member States is decreasing too slowly to meet the EU-targets. A new generation of measures is needed, underpinned by a new generation of research methods. Recent technology developments allow for this: naturalistic observations. This means that road user behaviour is observed unobtrusively in a natural setting for a longer period of time. This technique allows analysing the interrelationship between road user, vehicle, road and other traffic in normal situations, in conflict situations and in actual collisions. It will lead to a better understanding of these interrelationships.
The main objective of PROLOGUE was to prove the feasibility and usefulness of a large-scale European naturalistic observation study. The project aimed at road safety researchers and other stakeholders including car industry, insurance companies, driver training and certification organisations, road authorities, and governments. Whereas road safety is the main motive, the project also looked at the relevance for environmental issues, e.g. CO2 emissions, and traffic management.
Using the naturalistic observation research method will lead to a better understanding of road safety and help to realise an intrinsically safe road transport system, including in-car technology, self-explaining roads, driver training, etc.
Based on inventory studies, five small-scale field trials, and close involvement of user groups and stakeholders, PROLOGUE set out to result in recommendations and an outline for a large-scale naturalistic study, dealing with research questions, methodology and technology for data collection, data storage, data reduction, data mining and data analysis.
The five field trials were conducted in Israel, Austria, the Netherlands, Spain and Greece. The trials varied in a number of aspects, such as the technology used, type and number of cars involved, research questions addressed, target populations, samples and sampling strategies, data handling and storage, data reduction techniques and data analysis. From the technological point of view, all field trials included a technological system measuring basic g-force based driving parameters. Beyond this, the various technological systems varied substantially, ranging from simple accelerometers to fully equipped cars. This diversity in experiences contributes to the identification of aspects to be considered in a subsequent large-scale Naturalistic Driving (ND) study. Moreover, the trials provided an illustration of the type of information that ND research can provide.
Communication and dissemination to all potential stakeholders are vital to gain their support for and involvement in a large-scale European study.
The PROLOGUE consortium consisted of 9 partners institutes, well spread over Europe and included Israel. The consortium has a wide experience on all aspects relevant for naturalistic observations, and a large international network of road safety and transport experts.
Funding
Results
Conducting a large-scale ND study requires a great number of decisions and considerations, on all sorts of issues and at all levels. The findings and experiences in PROLOGUE finally resulted in a set of recommendations for a large-scale study covering the research questions, the potential users, methodology and equipment. The most important general recommendations are:
- The European Naturalistic Driving study should include pedestrians and (powered) two-wheelers (VRUs), and trucks, in addition to cars thus distinguishing it from the U.S. studies.
- An integrated data acquisition system is recommended because use of different technologies and vendors within the same project creates validation and data compatibility issues that lengthen the study and make it more expensive.
- Difficulties associated with recruiting drivers, as experienced in the SHRP2 project, should be taken into consideration when planning the large-scale study, and should be addressed in the design and the timetable of the study.
- In part of the study site-based and in-vehicle observations should be combined.
- Some specific research questions should be stated, and the design should be geared to answering them. An example of a design adaptation to specific research questions is over-sampling of certain groups, like young drivers, old drivers, or new vehicles.
- Automatic recording of behaviour should be supplemented by driver interviews e.g. to investigate look-but-did-not-see incidents with powered two-wheelers. The Naturalistic Driving database should also be enriched by adding other driver background data like sensation seeking, Driver Behaviour Questionnaire, and past violations and crashes.
- Emissions and on-line fuel consumption should be recorded for analysing ecodriving and environmental effects.
- Route and lane preferences and their relationship to background variables should be observed in order to provide relevant data for traffic management purposes.
- Inputs and/or insights from different stakeholders should be used to identify specific research questions.
- Cultural differences in driving patterns should be investigated; this requires data about type, number and locations for different observation sites.
- Some aspects of the data collection measures should be harmonized with those of SHRP2 and other large-scale naturalistic driving databases for the purpose of comparing European data with data from the U.S. and elsewhere and also for
Strategy targets
An efficient and integrated mobility system: Acting on transport safety (saving thousands of lives)