Reverse estimation of urban mobility patterns during pandemics using agent-based modeling

Abstract

In addressing pandemics like COVID-19, there is a crucial focus on proactive response research, predicting disease cases, and identifying risk areas. However, challenges arise due to limited human mobility data and methodological constraints in predicting travel patterns. To tackle this, our study introduces an Agent-Based Travel Scheduler (ABTS) model, simulating individual travel patterns using aggregated data sources. This model decomposes and forecasts travel behaviors by various criteria, such as age, weekdays/weekends, and trip purpose. The findings uncover varied travel behaviors across pandemic periods and demographic groups, highlighting complex movement patterns linked to infection risks. Moreover, the results show how different age groups adapt travel during pandemics, offering insights for targeted disease control strategies. By examining past pandemic-associated travel patterns, this study provides valuable insights for formulating effective proactive responses in future pandemics, guiding policy decisions to mitigate the spread of infectious diseases.