Driving simulators provide a safe, highly reproducible environment in which to assess driver behavior. Nevertheless, data reduction to standardized metrics can be time-consuming and cumbersome. Further, the validity of the results is challenged by inconsistent definitions of metrics, precluding comparison across studies and integration of data. No established tool has yet been made available and kept current for the systematic reduction of literature-derived safety metrics.
The long term goal of this work is to develop DriveLab, a set of widely applicable routines for reducing simulator data to expert-approved metrics. Since Matlab™ is so widely used in the research community, it was chosen as a suitable environment. This paper aims to serve as a case study of data reduction techniques and programming choices that were made for simulator analysis of a specific research project, the Simulated Driving Assessment. The initial set of Matlab™ routines was successfully tested by analyzing recognized metrics, such as Distance Headway, Time Headway, Time-To-Collision, and Reaction Time. The programming choices are described in this paper. This foundation of processes and Matlab™ routines will be used as the base for a DriveLab. The newly developed Matlab™ toolbox will be validated using the Realtime Technologies Incorporated (RTI) simulator.It will be generalized to other driving simulators so as to be a robust research tool for use by the scientific community.