Search the Noteworthy Practices database
This purpose of this case study is to describe Kentucky’s network screening methodology for all State-owned roads, as well as local roads classified as a collector street or above. The Kentucky Transportation Cabinet’s (KYTC) Highway Safety Improvement Program requires a data-driven process to identify sites with a potential safety need and prioritize projects. The KYTC partnered with the University of Kentucky’s Kentucky Transportation Center (KTC) to develop a network screening approach to prioritize locations statewide to be targeted for future safety improvement projects. This network screening approach addresses five focus areas: 1) Roadway Departure Corridors, 2) Cable Barrier, 3) High Friction Surface Treatment (HFST) Segments, 4) HFST Ramps, and 5) Intersections. The KTC analyzed statewide enterprise road, traffic, and crash data to develop safety performance functions (SPFs) that predict crashes on all facilities encompassed by each focus area. The KTC used cumulative residual (CURE) plots to assess SPF model performance and identify outliers or issues inherent in the dataset that lead to worse model fit. The CURE plot approach also underscores the importance of thoughtful and homogenous site segmentation for improved performance and meaningful network screening results. This network screening methodology applies a State-specific approach to rank locations based on higher-than-expected crashes and associated crash costs.