An Intelligent ABM-based Framework for Developing Pandemic-Resilient Urban Spaces in Post-COVID Smart Cities

Abstract

As of August 2022, the COVID-19 pandemic has accounted for over six million deaths globally. The urban population has been severely affected by this viral pandemic and the ensuing lockdowns, resulting in increased poverty and inequality, slowed economic growth, and a general decline in quality of life. This paper proposes a framework to evaluate the effects of the pandemic by combining agent-based simulations—based on Susceptible-Infectious-Recovered (SIR) model—with a hybrid neural network. A baseline agent-based model (ABM) incorporating various epidemiological parameters of a viral pandemic was developed, followed by an additional functional layer that integrates factors like agent mobility restrictions and isolation. It is inferred from the results that low population densities of agents and high restrictions on agent mobility could inhibit the rapid spread of the pandemic. This framework also envisages a hybrid neural network that combines the layers of convolutional neural network (CNN) and long-short-term memory (LSTM) architecture for predicting the spatiotemporal probability of infection spread using real-world pandemic data for future pandemics. This framework could aid designers, regulators, urban planners, and policymakers develop resilient, healthy, and sustainable urban spaces in post-COVID smart cities.