Data-Driven Approach for Measuring and Managing Physical Distancing in Subways during Pandemic Conditions

Abstract

Physical distancing was considered one of the key strategies for controlling the transmission of COVID-19. Its importance is particularly emphasized with regard to public transportation, on which passengers gather in confined spaces and create a high population density. Public transportation, although integral to the functioning of urban systems, possesses the potential to facilitate disease transmission. Its dual nature has motivated researchers and policymakers to propose transportation policies aimed at minimizing COVID-19 transmission. Nevertheless, quantitative assessments of physical distancing in relation to public transportation to support informed decision-making remain scarce. This study highlights the importance of data-driven analytics for understanding and assessing physical distancing when using public transportation. By applying a stochastic approach to smart card and train log data collected in Seoul, we estimated the routes traveled by subway passengers. We then calculated physical distances using the unit of links, which represents the flow of passengers between two consecutive stations, to obtain the spatiotemporal distribution of physical distances across the subway network. Using percentile ranks of these physical distances, we identified sections in which close physical distancing persisted after the pandemic. Additionally, we conducted an analysis comparing the travel time using subways with alternative modes of transportation such as buses and automobiles. Our data-driven analysis yields insights into areas within the public transportation system that require intervention and provides valuable information to policymakers for devising strategies for a safer and more sustainable public transportation system considering COVID-19 and potential future epidemics.