Improved Pedestrian Safety at Signalized Intersections Operating the Flashing Yellow Arrow

David Hurwitz, Oregon State University



Transportation facilities, when designed appropriately, attempt to provide a balance between safety and efficiency while acknowledging the design implications on their most vulnerable users. At signalized intersections, pedestrians are considered to be amongst the most vulnerable. When in the crosswalk at intersections without protected left-turn phasing, pedestrians are particularly at risk from left-turning vehicles. Though legally required to yield to opposing though vehicles and pedestrians until an acceptable gap is present, it is not uncommon for drivers to fail to observe pedestrians. Until recently, a wide variety of indications were in use across the US to indicate a permissive left-turn condition to the driver. In Oregon, the Flashing Yellow Arrow (FYA) has been used to indicate the permissive left-turn condition for approximately 10 years. With the addition of the FYA in the 2009 MUTCD, it is likely that its use will continue to increase nationally. Though many operational and safety issues have been studied about the FYA indication, this research proposes to fully investigate factors (e.g. opposing traffic volumes, pedestrian volumes, pedestrian expectation, intersection geometry, phasing sequence) that influence driver behavior in the context of the permissive left-turn conflict with pedestrians. To the best of our knowledge, this particular issue has not yet been studied in detail for the FYA. Specifically, the research seeks to study driver glance behavior to identify reasons why drivers are, "looking at but not seeing" pedestrians in or near the crosswalk or not searching for the presence of pedestrians at all. This research proposes a simulator-based approach. The research plans to first identify candidate FYA locations from historical crash data from the many installations in Oregon. From this candidate list, a selected set of intersections will be identified and modeled in the driver simulator. The simulator study will take place in the newly established Oregon State University Driving Simulator, a high fidelity one dimensional motion base driving simulator providing approximately 220 degrees of projection on three forward projection screens, one rear screen, and two LCD screens on the side view mirrors. Drivers will be exposed to approximately 20 independent left turn maneuvers during one 45 minute experimental trial. During each left turn maneuver, driver gaze information (location and duration) will be recorded. Additionally, vehicle trajectory and lateral position will also be collected. Different physical intersection characteristics such as number of opposing lanes, angle of skew, and presence of a median will be tested to determine if they influence the likelihood of pedestrian detection by the driver. Dynamic attributes such as volumes of opposing vehicular and pedestrian traffic will be varied as well to determine their influence on pedestrian detection. New signal control software available in Oregon allows the addition of a leading pedestrian interval prior to the permissive left-turn; we plan to study how this influences driver behavior. This experimentation will result in the identification of factors that increase the likelihood of pedestrian detection during PPLT phasing with the FYA. Recommendations will be proposed for what situations will warrant additional pedestrian protection, such as exclusive phasing.

Project Details

Project Type:
Project Status:
End Date:
December 31,2012
UTC Grant Cycle:
OTREC 2012
UTC Funding: