Research In Action

Understanding why crashes occur is key to preventing them. Most previous research has focused extensively on crashes that have already happened, but studying near crashes can provide additional data on driver errors that may lead to crashes and the types of evasive maneuvers drivers may use to avoid crashing.

I recently worked with a team of researchers led by Thomas Seacrist, MBE to analyze 4,818 near crashes from the Strategic Highway Research Program 2 (SHRP2) naturalistic driving database involving four age groups: teen (ages 16-19), young adult (ages 20-24), adult (ages 35 to 54), and older adult (ages 70 and older).The near crashes were classified into seven incident types: rear-end; road departure; intersection; head-on, side-swipe, pedestrian/cyclist, and animal. Near crash rates, secondary tasks, and evasive maneuvers were also compared across age groups.

Funded by the Center for Child Injury Prevention Studies (CChIPS), the study findings were recently published in the Journal of Safety Research and provide valuable insights into factors that can lead to crashes:

• Teen and young adult drivers had a higher rate of near crashes than adults and older adults, with the most occurring in rear-end and road departure scenarios.
• A longer time-to-collision at braking occurred during near crashes compared to crashes.
• Older adult drivers most commonly experienced near crashes at intersections.
• The type of evasive maneuver did not significantly vary across age groups.
• Timely execution of evasive maneuvers was a distinguishing factor between crashes and near crashes.

These findings can be used to develop more targeted driver training programs that focus on specific skill building and to help automobile manufacturers optimize advanced driver assistance systems (ADAS) technology to address the most common errors that can lead to crashes. Our team continues to find high value in the SHRP2 naturalistic program that provides access to video and numerical data for countless crashes and near crashes.This research is also leveraging experience from an NSF-funded big data and data science project in collaboration with Dr. Chris Yang and the Machine Learning team at Drexel University.