Naturalistic driving studies as a method to explore driving and driver behaviour in a natural environment have become increasingly popular in road safety research. However, data acquisition systems needed for these studies are expensive and require a profound technical expertise for the installation. This thesis reports on an alternative approach towards more affordable naturalistic driving studies – a smartphone-based system called Sensor Platform that leverages the phone’s sensors as well as external sensors to gather relevant driving data. In close cooperation with road safety experts, this project aimed to specify the requirements for such a system, develop a prototype, and evaluate from an user perspective as well as from a technical point of view. A focus group and an in-vehicle user study were conducted to gather the expert’s feedback. In order to judge the accuracy of Sensor Platform, a comparison to an industry-grade data acquisition system was performed on the real road. The analysis of the study data suggests that road safety experts like the high usability and value the time savings. Yet, in comparison to industry-grade data acquisition systems, Sensor Platform is not on par when it comes to data accuracy, mainly due to simpler filtering algorithms. All in all, the thesis adds to the knowledge of mobile data acquisition systems while also providing a basis for future road safety applications such as real-time interventions.
展开▼
