A CRITICAL ANALYSIS OF ROADWAY AND CRASH ATTRIBUTES, INCLUDING BALL BANK INDICATOR, IN PREDICTION OF CURVE LANE DEPARTURES
Knezevich, Ronald William
MetadataShow full item record
More than 25 percent of roadway fatalities are associated with a horizontal curve, and the vast majority of these crashes are due to lane departures. To mitigate these fatalities, engineers often utilize safety performance functions (SPFs) or crash prediction models to critically assess the roadway environment. The goal of this thesis is to assess influential crash types and explanatory variables in the prediction of curve lane departures. To accomplish this goal, there are three objectives. 1) Identify if ball bank indicator (BBI), a measure of lateral forces, should be used in SPFs for curve lane departures. BBI values gauge both the roadway geometry and driver behavior. 2) Assess the implications of selecting specific crash types for crash prediction models. 3) Leverage the outcomes of this thesis for practical use. A case study on rural curves in GDOT districts 1, 2, and 6 is conducted, and there are 3 main outcomes.1) An SPF had better performance when BBI was included as an explanatory variable. 2) Using an SPF developed with only curve lane departures allows one to establish more focus when conducting safety assessments than using a SPF with all crash types (i.e., rear ends, angle, etc.). 3) A SPF is presented in a format intended for practical use. In future research it is recommended to 1) utilize BBI especially if able to be crowdsourced, in prediction of curve lane departures, and 2) develop focused SPFs with specific (rather than all) crash types such as curve lane departures.