01519668

Component

Using second-by-second naturalistic data collected by a large fleet of global positioning system (GPS) instrumented vehicles, a relationship among longitudinally-observed driving behavior and exposure metrics (mileage, driving duration, speeding, hard acceleration or deceleration, cruise speed duration, unfamiliar roadway exposure, left/right turn exposure, and crash location exposure) and numbers of crash involvements was investigated. With the initially proposed 413 driving behavior exposure metrics, this study identified 22 driving behavior exposure metrics that demonstrated significant difference between the crash-involved and the crash-not-involved drivers after the statistical analyses. Further, accompanied by the 22 driving behavior exposure metrics, the study determined five most contributing metrics in predicting the potential crash involvement status of individuals as the following: mileage in rural area during the afternoon, frequency of over-speed activity (>= 15mph) per mile on arterials during the morning, frequency of hard deceleration (>= 6mph/s) per mile on local streets during the night, frequency of hard deceleration (>= 8mph/s) per mile on freeways during the morning, and frequency of hard deceleration (>= 8mph/s) per mile on freeways during the afternoon. Such GPS-observed driving behavior exposure metrics supported existing safety campaigns such as anti-aggressive driving and speed enforcement to reduce the risk and the number of future crashes. When combined with the automobile insurance premium structure, these driving behavior and exposure metrics could be used to help drivers rectify their driving habits and patterns through proper education and monitoring programs. Immediate implantation of such driving metrics as part of safety measures is arduous due to requirement of sophisticated and expensive data collection systems. However, the recent advancement in mobile telecommunication technology would provide opportunities to monitor individual driving activities for a long-time period.

This paper was sponsored by TRB committee ANB20 Safety Data, Analysis and Evaluation.

01503729

14-4232

English

12p

2014

Transportation Research Board 93rd Annual Meeting

Digital/other

References; Tables

Automatic data collection systems; Behavior; Crash exposure; Crash risk forecasting; Drivers; Global Positioning System; Metrics (Quantitative assessment); Periods of the day; Speed; Traffic safety; Trip length

Data and Information Technology; Highways; Safety and Human Factors; I83: Accidents and the Human Factor

Transportation Research Board Annual Meeting 2014 Paper #14-4232

TRIS, TRB, ATRI

Jan 27 2014 3:27PM