Rural Speed Safety Project for USDOT SDI






Subasish Das and Kay Fitzpatrick

December 13, 2018


Outline

  • Overview
  • Data Preparation
  • Exploratory Data Analysis
  • Statistical Models
  • Decision Support Tool
  • Next Steps

Overview

  • Objective
    • Integrate vehicle travel speed data with geometric and traffic data to identify the relationship of speed and crashes
  • Product
    • Create a protype Decision Support Tool based on expected crashes

Data Preparation

  • Two States: Washington, and Ohio
  • Three Databases
    • Highway Safety Information System (HSIS)
    • Highway Performance Monitoring System (HPMS)
    • National Performance Management Research Data Set (NPMRDS)
  • Year: 2015

Questions Identified

  • Q1: Do different speed measures contribute to crash outcomes?

  • Q2: Is there more variability in speeds just prior to a crash?

Options


  • Year
    • 2015
  • County
    • All counties
    • Select a county
  • Faciltity Type
    • Rural two lane
    • Rural Interstate
    • Rural multilane divided
    • Rural multilane undivided
    • Rural others
  • Severity
    • All
    • Fatal and Injury

Decision Support Tool

App link: https://ldwhite.shinyapps.io/RuralSpdSafety_Demo/

Answer to Questions

  • Q1: Do different speed measures contribute to crash outcomes?
    • Yes!
    • Developed a prototype of Decision Support Tool to illustrate number of expected crashes on segments
  • Q2: Is there more variability in speeds just prior to a crash?
    • Yes!
    • Developed models that use travel speed to classify crash and non-crash event

Next Steps

  • Decision Support Tool
    • Refine the models for data structure 2
    • Apply daily/hourly level models in the prototype
  • Crash/ No Crash
    • Apply machine learning models (for example, support vector machine) to reduce classification errors

Questions?