Understanding Crashes Involving Roadway Objects with SHRP 2 Naturalistic Driving Study Data

This project used the second Strategic Highway Research Program (SHRP 2) naturalistic driving study (NDS) data as an alternate data source to police-reported crash data to better understand roadway object crashes. The objectives included determining crash causation, recommending strategies for crash prevention, and understanding the implications for highly automated vehicles (HAVs). Researchers addressed these objectives with a three-pronged approach: (1) a detailed engineering study of roadway object events to identify and quantify effects of a large number of relevant variables; (2) a machine-vision-oriented study to document the implications of roadway object events on machine vision performance; and (3) a detailed case study analysis of representative roadway object events to provide further qualitative results on how and why roadway object crashes occur and what potential actions can prevent such events effectively.