“Use of Big Data to REDUCE Traffic Crashes via Enhancements to Models Using Surrogate Measures of Risk”

• Activity Number: 220 - Contributed Poster Presentations: Transportation Statistics Interest Group
• Type: Contributed
• Date/Time: Tuesday, August 4, 2020 : 10:00 AM to 2:00 PM
• Sponsor: Transportation Statistics Interest Group
• Author(s): John McFadden* and Seri Park and Q.B. Chung and Roya Amjadi and Andrew Graettinger
• Companies: Federal Highway Administration and Villanova University and Villanova University and Federal Highway Administration and University of Wisconsin - Milwaukee
• Keywords: Big Data; Reduce; Trafffic crashes; Save Lives; Crash Prediction Model; Enhancements

Abstract

JSM 2020 ABSTRACT SUBMISSION - EVERYONE COUNTS-Data for the Public Good

“Use of Big Data to REDUCE Traffic Crashes Via Enhancements to Models Using Surrogate Measures of Risk”

By: John McFadden1, Q.B Chung2, Roya Amjadi3 and Seri Park4

FHWA Every Day Counts (EDC) Data Driven Safety Analysis (DDSA) promotes advancments in highway safety analysis to provide information for effective investment decisions on the nation’s highway system.

Quantitative safety analysis in the HSM includes crash prediction models which use traffic volume and roadway features to predict expected value of crashes per year for a site. Every traffic crash includes three (3) elements: the driver, the vehicle and the roadway/traffic variables. This research proposes an alternative approach that uses surrogate risk based variables into crash prediction models to better explain the variability in traffic crashes.

This project is a proof of concept incorporating of driver related risk surrogate measures to predict traffic crashes. The research links the human element of traffic crashes via regional and local socio-economic characteristics referenced in a geo-spatial manner to existing crash models.