Project

During the past decade, University of Iowa and Lytx (former DriveCam) have conducted research on coaching programs for teenage drivers. As a result of this work, a large database of real-world crashes (n~=6000) has been collected. An initial analysis of a subset of these crashes has recently been conducted by. U. Iowa in a project sponsored by the AAM Foundation. In parallel, Chalmers and AB Volvo have run the project ANNEXT together with Lytx, where 133 crashes and 60 near event-triggered crashes have been analyzed. This data was obtained through Lytx’ standard safety management program for commercial fleets, including trucks, buses and cars. The ANNEXT project also involved the development of methods and tools for analyzing event-triggered naturalistic driving data, which have also been carried over to, and further developed, in the SHRP2 S08 analysis project as well as the BBS China project, which involves collection of Lytx data in China. The work has so far resulted in two conference papers (Bärgman et al., 2013) and Engström et al. (2013), and further publications are under way. These initiatives are the worlds’ first major efforts on detailed crash investigations based on event-triggered naturalistic driving data (today, only Lytx and their main competitor SmartDrive have access to such data). In order to maintain their leading position in the field, University of Iowa, Lytx, Chalmers, Volvo Cars and AB Volvo have decided to join forces to continue these analysis and methods/tools development efforts. The first step, which represents the scope of the present proposal, will be combine the analysis methods and tools developed to date by U. Iowa and Chalmers/Volvo and apply these to a subset of the teen driver data. The work thus includes: developing a code book suitable for addressing 3 key research questions (Run off road crashes, Intersection crashes, and Lead vehicle crashes and near-crashes); conducting initial coding and analysis of Lytx data to assess the code book questions and develop further research projects; and establish data processing tools to extract kinematic data from Lytx videos. The driver actions prior to crash will also be analysed as input to injury causation research (e.g. research related to drivers being in “bad” positions/posture during the crash). It may also be that other roads users’ behavior will be modeled separately. The initial focus of the SAFER/UoI research will be on lead-vehicle, intersection and run-off-road conflicts, but later also other conflicts types will likely be included.