Using smart card data to model public transport user profiles in light of the COVID-19 pandemic

The COVID-19 pandemic caused an unprecedented impact on public transport demand. Even though several studies have investigated the change in the use of public transport during the pandemic, most existing studies where large passive datasets have been considered focus on the drop in ridership at the aggregate level. To address this gap, this research aims to identify and model profiles of passengers considering their public transport recovery after the long-term lockdown in Santiago, Chile, during the early stage of the pandemic. The methodology proposed a three-stage approach associated with the analysis of smart card records. First, cardholder residential areas were identified to enrich the available data by integrating demographic information from the census. Then, a clustering analysis was applied to recognise distinctive classes of users based on their public transport usage change between the pre-pandemic and the post-lockdown phase. Finally, two different models were implemented to uncover the relationships between class membership and travellers’ characteristics (i.e. travel history and demographic characteristics of their residential area). Results revealed a heterogeneous recovery of public transport usage among passengers, summarising them into two recognisable classes: those who mainly returned to their pre-pandemic patterns and those who adapted their mobility profiles. A statistically significant association of travel history with the mobility adaptation profile was found, as well as with aggregate socio-demographic attributes. These insights about the extent of heterogeneity and its drivers can help in the formulation of specific policies associated with public transport supply in the post-pandemic era.