Center for Injury Research and Prevention

Understanding Driver Attention Through Machine Learning and Big Data ​​​​​​​

January 17, 2019

There has been a large effort in the last few years to better understand driver behavior. Traffic safety researchers like myself pay close attention year after year to the motor vehicle crash-related fatality reports released by the National Highway Traffic Safety Administration. While these numbers generally go down, an alarming uptick was observed in 2015 and 2016. One of the major causes for this concerning increase in crashes was driver inattention.

Being fully engaged in the driving process, especially for inexperienced teen drivers, is crucial in keeping our roads safe. To help prevent inattention-related crashes from occurring, colleagues from Drexel University and the Virginia Tech Transportation Institute (VTTI) and I have been awarded a $1 million Big Data and Data Science grant from the National Science Foundation (NSF) to develop predictive analytics techniques of engagement in the driving task.

Led by Christopher Yang, PhD, of Drexel’s College of Computing & Informatics, the research team includes Santiago Ontañón, PhD, also of Drexel, and Charlie Klauer, PhD of VTTI. We are excited to extend the capability of machine learning, sensor informatics, and driver behavior analytics to recognize and predict driver disengagement so that we can develop training to maintain a high level of attention while driving.

With advanced driver assistance systems and self-driving capabilities becoming more and more available to drivers, it’s all the more important to remind drivers that they are still in charge and to maintain their focus on the road whether or not automated features are engaged.

Over the course of the study, we will be processing the large volume of diverse factors collected as part of the federal government’s Strategic Highway Research Program (SHRP2) Naturalistic Driving Study to develop predictive analytics algorithms.

I look forward to once again tapping into the SHRP2 database to better understand factors that lead to crashes. My work with CIRP colleague Thomas Seacrist, MBE has allowed us to scrutinize individual crashes with a focus on teen driving, and I am deeply appreciative to the Center for Child Injury Prevention Studies (CCHIPS), which supported the launch of this research.

The National Science Foundation should also be commended for developing its Big Data and Data Science grant program to encourage novel approaches in computer science, statistics, computational science and mathematics, as well as the development of innovative applications in domain science. As part of NSF’s focus on education and training, we plan to incorporate the research findings in courses offered as part of the Master of Science program in Health Informatics at Drexel University and to organize workshops, conferences, and seminars to disseminate the research outcomes.

**Like what you’ve read? Subscribe to Research in Action to have the latest in child injury prevention delivered to your inbox.**