A Dynamic Model Integrating Passenger Satisfaction Analytics and Pedestrian Simulation for Optimizing Metro Station Performance and Enhancing Passenger Experience

Abstract

This study develops a fully dynamic model to systematically assess metro station performance and passenger satisfaction, integrating the American Customer Satisfaction Index (ACSI) model with Structural Equation Modelling (SEM), Partial Least Squares (PLS) regression, and agent-based pedestrian simulation. A conceptual framework, consisting of 28 key indicators for SEM, refines passenger satisfaction indicators (PSI) through PLS regression to explore relationships between satisfaction drivers and metro performance. Importance-Performance Analysis (IPA) is also employed to identify key satisfaction factors. The research uniquely integrates SEM-PLS results with pedestrian simulation. ACSI SEM-PLS provides insights into passenger satisfaction, which are then used in ped simulation to predict the impact of projected ridership growth (3%, 5%, and 7% from 2030-2040) on station key performance indicators (KPIs). This combined approach offers a full understanding of passenger perceptions and operational dynamics. The study finds that current infrastructure will not support the anticipated ridership increase, leading to declines in station KPIs in efficiency (Space Density SD) and passenger experience (Passenger Density PS, Passenger Speed PS, Travel Time TT and Passenger Dissatisfaction Level PDL). To address these gapes, the study presents URBM model, which integrates improvements like expanded ticketing, improved circulation, and other strategies. These enhancements lead to significant gains across various KPIs metrics: SD improves by 44-56%, PD increases by 70-80%, PS rises by 53-70%, TT decreases by 11-16%, and PDL is reduced by 37-41%. The research contributes to knowledge by developing a KPI-driven framework that synthesizes satisfaction analytics with simulation modelling, thereby advancing methodological approaches in metro station design.