Concept to Countermeasures

UNR Selected for Urban Pedestrian Crash Research

Traffic safety professionals know that over 90% of the time the driver is either completely or partially at fault in fatal crashes, but how do we address this problem?
Photo: Head UnitThe Second Strategic Highway Research Program (SHRP 2) set out to investigate the problem by outfitting volunteers’ vehicles with video monitoring, Global Positioning System (GPS), radar, and a host of other data collection equipment. The goal was to get a better understanding of how drivers behave behind the wheel and use that knowledge to improve both roadway infrastructure and vehicle technology and performance. Over a 6-year study period, the SHRP 2 Naturalistic Driving Study (NDS) enlisted 3,147 driving volunteers from six locations across the United States to collect 5.4 million trips, 3,958 vehicle-years, and 49.6 million vehicle-miles of data. Additional information was collected on roads most traveled by the volunteers.

To utilize the large amount of data, SHRP 2 partners AASHTO (American Association of State Highway Officials) and FHWA (Federal Highway Administration) initiated Concept to Countermeasures – Research to Deployment Using the SHRP 2 Safety Data, and a national request for projects resulted in 11 research projects.

Photo: Machine vision algorithmOne of the projects, led by Dr. Hao Xu from the University of Nevada, Reno (UNR) focuses on how drivers negotiate turning movements at urban intersections. The long term goal is to identify infrastructure improvements to reduce pedestrian crashes at urban intersections. Six intersections (4- or 6-lane roadways) that experienced pedestrian crashes were selected: four in Washington State and two in Florida.

To maintain the anonymity of the driver volunteers, the team from UNR traveled to Virginia Tech to view secure facial videos. The team reviewed video of the drivers’ faces and recorded driver’s glances before, during, and after the turn, in addition to other valuable data elements. The team compiled and analyzed both the vehicle sensor data and driver behavior data.Photo of Intersection

Dr. Hao Xu and Chuck Reider presented the research findings to the AASHTO Safety Task Force and FHWA in Arlington, Virginia in October 2015. Some interesting observations of right turn drivers include:

  • The number of right windshield glances is affected by conflicting traffic. Low flow (less than 1,000 vehicles per hour (vph)) and high flow (greater than 2,000 vph) resulted in fewer glances to the right where pedestrians may be crossing.
  • “Right Turn On Red Must Yield to U-Turn” sign had the highest correlation with pedestrian crashes, six times higher than any other sign. That same intersection had the fewest right windshield glances per trip, only about 1 in 10 trips.
  • The intersection with “Turning Vehicles Yield to Pedestrians” produced 30% more right windshield glances than the intersection without the signs.

The State of Nevada is aware of the seriousness of pedestrian crashes and this project will provide clues on how to make urban intersections safer for pedestrians. One more step towards Zero!