ASSESSING THE INFLUENCE OF MANAGED LANE SEPARATION TYPE ON LANE DEVIATION

Abstract: Managed lanes (MLs) are widely adopted by transportation agencies to enhance mobility and improve highway safety. These lanes are intentionally separated from general-purpose lanes (GPLs) to provide controlled and optimized use of the ML facilities. By maintaining this separation, transportation agencies can implement targeted strategies to alleviate congestion, improve overall traffic flow, and achieve other intended objectives. Managed lanes (MLs) are typically constructed adjacent to general-purpose lanes and are separated by various types of barriers. The separator types considered in this study include pylons, concrete barriers, buffers, and wide buffers (medians). Driver behavior, such as lane deviation, is influenced by the type of separator used. Lane deviation, defined as the offset between the position of the vehicle's centroid and the centerline of the lane, is a key safety performance measure, as drivers encroaching on adjacent lanes can compromise roadway safety.

This study utilizes naturalistic driving data (NDS) from 8,051 events recorded in the Second Strategic Highway Research Program (SHRP2) dataset from the state of Washington to analyze the impact of different managed lane separator types on lane deviation. The dataset includes time-series parameters for each event, such as the variable “lane distance off center,” which corresponds to lane deviation. Additional data from SHRP2 includes anonymous driver and vehicle identifiers, enabling the retrieval of demographic and vehicle characteristics. Traffic characteristics, such as the average annual daily traffic (AADT) for each section, along with geometric characteristics like the number of lanes, lane width, and shoulder width, were obtained from the Washington Department of Transportation (WSDOT) Geospatial Open Data Portal.

Preliminary results from an Analysis of Variance (ANOVA) indicate a significant difference in mean lane deviation across these separator types. Notably, all four separator types exhibited net negative lane deviations, meaning that drivers tend to steer away from the separator when driving in the inside lane of a managed lane. A mixed-effects multinomial model was employed to assess the influence of various explanatory factors, including separator type, geometric characteristics (e.g., number of managed and general-purpose lanes, shoulder surface, shoulder width, lane width), demographic factors (e.g., age, gender, ethnicity), and vehicle characteristics (e.g., vehicle classification). The findings offer insights into how separator type and other variables affect lane deviation, with implications for roadway design and safety strategies.