Home About ATSIP Program Awards Venue Sponsors/Exhibitors
 

LINEAR RISK TERRAIN PREDICTING IMPAIRED DRIVING

 

Abstract: This presentation will cover the progress of the development of a linear risk terrain model to identify road segments that are at an elevated risk for impaired (alcohol and drug) driving, so that these highways can be targeted by law enforcement for saturation patrols. This presentation will cover the results after the preliminary stages of testing the linear risk terrain model.

This model will show the progression and the accuracy of the linear risk terrain model from the original model of using only three datasets, to the improved model of factoring in more datasets. The model will include various sources of feature input data: citations, crashes, liquor licenses, place of last drink, driver’s destination prior to being stopped, weather, population, and more. By aggregating these linear data layers together, we are able to produce an overall risk value for each road segment included in the analysis. This model will test to verify if using place of last drink is a more accurate source to use for the predictive linear risk terrain model for impaired driving, as well as testing what other factors, such as demographics or weather, have a greater impact in prediction of impaired crashes.


Alicia Shipley - GIS Analyst, Washington College

 

 


logo
Quick Links

Home
ATSIP
Contact

Chris Osbourn

forum@atsip.org
Connect


fb_logo   linkedin