THESIS ABSTRACT

Abstract

The rapid urbanization of Kathmandu Valley, coupled with increasing vehicle and pedestrian traffic, has highlighted the critical need for pedestrian safety, particularly at intersections. Despite the growing number of vehicles and pedestrians, pedestrian safety remains a largely overlooked aspect of traffic management. The Machhapokhari intersection, an unsignalized and uncontrolled junction, has been identified as particularly dangerous for pedestrians, with high crash frequencies and pedestrian fatalities. This study focuses on the assessment of pedestrian safety by the Ordered Logit (OL) model at the Machhapokhari intersection and developing a comprehensive model by applying the Analytical Hierarchy Process (AHP) and Multiple Linear Regression (MLR) analysis at crosswalks of the same intersection. Pedestrian safety assessment was conducted using a structured questionnaire survey of 400 pedestrians crossing the crosswalks of the Machhapokhari intersection. The survey data were analyzed using an ordered logistic regression (OL) model to identify the qualitative factors influencing pedestrian safety perceptions. Key variables included pedestrian age, previous crash history, vehicle volume, road width, traffic control, and pedestrian behavior. The study also developed a Pedestrian Safety Index (PSI) model based on AHP results, providing an objective measure of safety at the intersection. Findings from the study indicate that younger pedestrians (age=15-24 years) and those controlled by traffic perceive higher safety levels. Factors such as traffic police control, pedestrian road markings, and less road width at crossings of pedestrians significantly improved safety of pedestrians at the Machhapokhari intersection. Conversely, higher vehicle volume and speed were associated with lower safety perceptions. The PSI model validation confirmed its effectiveness, with a mean absolute percentage error (MAPE) of 2.76%, indicating high predictive accuracy. The study concludes that pedestrian safety at the Machhapokhari intersection is generally low, with most respondents perceiving it as highly unsafe. The research offers actionable insights for researchers and policymakers to study pedestrian safety at similar intersections across Kathmandu and other urban areas.

Keywords

Pedestrian Safety, Analytical Hierarchy Process (AHP), Multiple Linear Regression (MLR), Pedestrian Safety Index (PSI), Ordered Logistic Regression, Traffic Control.