Analysis of Driver Behavioral Adaptation to the Lateral Drift Warning System

Open Access
Author:
Greenstein, Adam Lawrence
Graduate Program:
Civil Engineering
Degree:
Master of Science
Document Type:
Master Thesis
Date of Defense:
October 28, 2009
Committee Members:
  • Paul Peter Jovanis, Thesis Advisor
Keywords:
  • driver
  • behavior
  • adaptation
  • safety
  • roadway departure
  • crash
Abstract:
The desired effect of in-vehicle intelligent transportation system (ITS) safety devices is to reduce the likelihood of crash occurrence. The Roadway Departure Crash Warning (RDCW) alert system, tested by the University of Michigan Transportation Research Institute (UMTRI) in 2004, was installed in test vehicles to determine if drivers changed their behavior over a period of four weeks based on the presence of roadway departure alerts. This analysis of adaptation while driving behind the wheel of an instrumented vehicle considered the effects of various driver characteristics on adaptation trends, including gender, years of driving experience, personal habits, and predispositions. Assessing changes in alert frequency over time is a direct way to determine if certain forms of adaptation exist while both ignoring and including driver descriptors. If alert frequency decreases over time, one can say that the system improved safety for drivers. The major findings showed that the two following statements are true: Smokers decrease alert frequency over time less substantially than do non-smokers; males decrease alert frequency over time more substantially than do females, and their adaptation is more consistent and more significant. Higher levels of sensation-seeking desires result in less substantial decreases in alert frequency over time, while higher risk perception levels indicate more substantial decreases. Exposure, an important factor in analyzing event frequency, was shown to significantly positively correlate with alert likelihood. However, differing levels of driving experience and locus of control do not show clearly-definable adaptation trends. Some results may have been affected by drivers’ abilities to properly use technology or issues surrounding the relatively small sample size used in the study, both of which could be accounted for in future research of this type.