Award Date


Degree Type


Degree Name

Master of Science in Engineering (MSE)


Civil and Environmental Engineering and Construction

First Committee Member

Shashi Nambisan

Second Committee Member

Hualiang Teng

Third Committee Member

Mohamed Kaseko

Fourth Committee Member

Ashok Singh


This research developed a methodology to evaluate safety proximate to transit stops through a Safety Index, identified key factors associated with fatal crashes proximate to bus transit stops, and illustrated the applications of these using a case study based on the Las Vegas metropolitan area in Nevada, USA. The methodology is based on motor vehicle crashes that involve pedestrians recorded in the proximity of transit stops, and the severity of the crash outcomes. It focused on crashes which involved pedestrians that occurred within walking distances of individual transit stops. The severity of the crash outcomes was classified based on the KABCO scale (K = fatal, A = incapacitating injury, B = non-incapacitating, C = possible injury, and O = no injury/property damage), a classification scheme that is commonly used for such analyses. The KABCO values used in this research are those recorded in individual crash reports. Further, crashes within each of the KABCO categories were weighted based on their respective Value of Statistical Life (VSL), a technique that is used commonly by the US Department of Transportation for Benefit-Cost Analyses. From a review of the literature, walking distances (accessible areas) proximate to transit stops were estimated based on walking speeds of 3.5 feet per second, and typical times that individuals spend walking to or from transit stops of 5 minutes and 10 minutes. The Safety Index is based on a weighted count of crashes recorded in these zones. The weightages were based on VSL values for each of the KABCO classes, and also inversely based on the maximum walking distance for the accessible areas.Using the SI for each stop, a three-tiered system was developed to categorize the stops based on their relative risk: low, medium, high. Using data recorded in individual crash reports, a binary logistic regression model was developed to identify key factors associated with fatal crashes. ArcGIS, a Geographic Information System (GIS) software was used to facilitate data capture, analyses, and visualization. For the case study, the data sets used were street centerline files along with network attributes, fixed-route transit route and stop location data, and crash data with their attributes - all for the Las Vegas Metropolitan area in Clark County, Nevada. The analyses used kernel density estimation, and spatial autocorrelation techniques, which are built-in spatial analyses tools available in ArcGIS. The results of the various analyses are presented graphically using GIS exhibits and in tabular format. The results of spatial autocorrelation methods indicated that the crashes were not randomly distributed. Angle and rear-end crashes were the predominant crash types identified in the vicinity of transit stops. Of the 3,364 transit stops analyzed, 25 percent of stops were characterized as low-risk, 25 percent under medium-risk, and 50 percent under high-risk category. Results from the binary logistic regression model indicated that pedestrians have higher survival rates during daytime crashes, and that driver factors such as “exceeded authorized speed limit” had a significant impact on the crash severity. Fatal crashes were more significant when pedestrians had “dark clothing,” “lying and/or illegally on the roadway,” and for “improper road crossing.”


Pedestrian Safety; Safety Index; Transit Stops; Walking distance


Civil Engineering | Transportation

File Format


File Size

8100 KB

Degree Grantor

University of Nevada, Las Vegas




IN COPYRIGHT. For more information about this rights statement, please visit