Currently there is not an effective way to evaluate intersection designs and compare alternatives for bicycle and pedestrian quality of service. This problem is not unique to NCDOT, and, in fact, the National Highway Cooperative Research Program (NCHRP) funded project 07-25 to develop a methodology to allow for the comparison of expected safety, operational, and comfort performance of alternative intersection and interchanges (A.I.I.s) with conventional intersection alternatives.
The NCHRP 07-25 research team developed a “20-Flags" method as a sketch-level evaluation tool to specifically assess bicycle and pedestrian safety of an intersection design within a larger intersection evaluation control (ICE) process. This method identifies three pedestrian-only flags, seven bicycle-only flags, and 10 bicycle- and pedestrian-related flags that correspond to 15 different measures of effectiveness to determine the performance of a given intersection. The flags and associated measures identified were based on some data and expert input, but most thresholds were adapted from prior research and guidance documents like the Highway Capacity Manual (HCM) and the AASHTO Green Book. Since the 20-Flags method is not evidence-based in empirical crash data or associated crash patterns, there is a need to determine the usefulness of this method to adequately assess intersection design for safety performance. Research is needed to confirm that reducing the number of elements flagged for a given intersection design will result in reducing bicyclist and/or pedestrian crashes. If a positive corollary relationship exists, and can be demonstrated through research, between the flags and crash outcomes, the 07-25's method has the potential to be utilized to compare intersection designs that are safe for pedestrian and bicyclist activity.
The objective of the research proposed is to test the hypothesis that crash data and flags will generally trend in the same direction. In other words, intersections with more elements flagged will reflect higher
rates of bicycle and/or pedestrian crashes than intersections with fewer flags. To do this, we anticipate collecting the data needed to apply the 20-Flags method to 300 intersections across North Carolina and compare the index of resultant flags to the crash data at the same locations over the past five to 10 years.
