IAP Project: Pedestrian Safety and High Visibility Markings

This research study was sponsored by USDOT/AASHTO with money flowing through CUBRC. The study proposed to assess the safety benefits of various High Visibility Crosswalk (HVC) designs and implementations. The dataset delivered consists of a sample of time series variables collected as part of the SHRP2 Naturalistic Driving Study. Time series data (de-identified) were requested from traversals of six HVCs in Buffalo, NY. 1,575 traversals of the intersection are included, representing 85 drivers. The HVCs were installed during the data collection period. For comparison, the sample of traversals include traversals of the HVC area before and after the installation of the HVC. The times series data include: vehicle trajectory data (e.g., GPS coordinates, acceleration, and velocity), other vehicle performance data (e.g., brake pedal activation, steering wheel position), and video of the forward roadway. The time series data for each HVC traversal covers the portion of time the driver is on the two roadway links constituting each traversal. Time series data exclude data near the beginning and end of a trip (defined as a pre-determined distance from a trip origin or destination). The distance contains a limited random noise element to further anonymous the trip.