Publication

Road accidents in the EU annually account for over 40,000 deaths. More than 90% of these road accidents are caused by driver impairment. Many on-the-market intelligent vehicle systems (IVSs) promise to reduce road accidents by supporting the driver in unpredicted and hazardous situations. However, the extent to which IVSs such as frontal collision warning and lane departure warning can prevent accidents is still uncertain. Large scale field-operational-tests (FOTs) aim at evaluating the actual impact of such IVSs in real traffic by collecting huge amount of data. EuroFOT is one of the first of these FOTs in Europe and committed to collect data from 1500 customer-vehicles for one year. Together with the potential of objectively evaluate IVSs in real traffic conditions, FOTs studies also present a number of new challenges intrinsically related to 1) the complexity of studying a real traffic environment and 2) the large scale of such a study. More specifically, the quasi-experimental conditions of the data collection and the need to harmonize and coordinate the study across very wide geographical areas require extra efforts through all the FOT study. In euroFOT – one of the first large-scale FOTs in Europe – these challenges have been faced, so far, in relation to the definition of the 1) IVSs under test and 2) hypotheses to be tested in the project. Specifically, the In-Vehicle Systems for Driving Support subproject addressed these new challenges by developing and following common procedures for the 1) description of the 8 different IVSs under evaluation, 2) formulation and harmonization of hypotheses across the 9 OEMs participating in the study, 3) combination of systems tested at different OEMs, and 4) specification of hypotheses to be tested. By following these procedures, specified hypotheses keeping into account a quasi-experimental set-up have been formulated, ranked, and harmonized across 9 OEMs and 28 partners in the project. In this paper, feasibility of hypotheses testing in the context of the FOT challenges is also discussed.