Date of Award
Master of Science
Civil, Construction, and Environmental Engineering
Peter T. Savolainen
Distracted driving has become a severe threat to traffic safety due in large part to the proliferation of in-vehicle smart technologies, the ubiquity of cell phones, and a general societal shift towards constant mobility and connectivity. Research has consistently demonstrated adverse consequences to engaging in a distracting secondary behavior while operating a motor vehicle. Much of the prior research in this area has leveraged data from traffic simulators and police-reported crash data, resulting in estimates as to the impacts of distraction on crash risk. However, research has been more limited under actual driving conditions and there remain important gaps with respect to how distracted driving and the associated crash risks vary across drivers and roadway environments.
This study addresses this gap by utilizing disaggregate time-series data from the second Strategic Highway Research Program (SHRP2) Naturalistic Driving Study (NDS) to conduct an in-depth investigation of various safety-focused aspects of distracted driving. The high resolution data were provided at 10 Hz resolution through a series of cameras and mechanical sensors. These operational data were integrated with geometric information from the companion Roadway Information Database (RID), as well as with data related to driver behavioral characteristics, risk perceptions, and risk-taking behavior from a series of participant surveys. Collectively, these sources resulted in a robust dataset of vehicle, roadway, weather, and driver behavioral parameters.
Various aspects of distracted driving were investigated as a part of this analysis, including the effects of distraction on driving performance. More specifically, the effects of various types of distraction on driver speed selection behavior was examined. Additional analyses assessed how the prevalence of various types of distracting behaviors varied based upon driver characteristics, roadway geometry, traffic conditions, and environmental conditions. As a part of these investigations, a series of random effects linear and logistic regression models were estimated with the disaggregate time-series information. Risk models were also estimated to determine how various types of distractions impacted the likelihood of a crash or near-crash event. Ultimately, the results suggest that drivers generally adapt their behavior based upon the level of risk posed by various driving environments. These environmental factors, along with various driver-specific factors, were shown to influence speed selection, as well as proclivity for participating in various types of distracting behaviors. In turn, these distractions were found to exacerbate crash risks, with marked differences exhibited based upon the degree to which the distracting behaviors required drivers to direct their attention away from the primary driving task.
Trevor Joseph Kirsch
Kirsch, Trevor Joseph, "An analysis of the crash risk and likelihood of engaging in a distraction while driving using naturalistic, time-series data" (2018). Graduate Theses and Dissertations. 16392.