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

Area of Science:

• Epidemiology
• Urban Planning
• Computational Modeling

Background:

• The COVID-19 pandemic caused over six million global deaths by August 2022.
• Urban populations faced severe impacts including poverty, inequality, and reduced quality of life due to the pandemic and lockdowns.
• Existing models require enhancement to predict and mitigate pandemic effects in urban environments.

Purpose of the Study:

• To propose a framework combining agent-based simulations and hybrid neural networks for evaluating pandemic effects.
• To assess the impact of urban factors like population density and mobility restrictions on pandemic spread.
• To develop a predictive model for infection spread in future pandemics.

Main Methods:

• Developed a baseline agent-based model (ABM) using the Susceptible-Infectious-Recovered (SIR) model with epidemiological parameters.
• Integrated a functional layer into the ABM to incorporate agent mobility restrictions and isolation.
• Envisioned a hybrid neural network combining convolutional neural network (CNN) and long-short-term memory (LSTM) for spatiotemporal infection spread prediction.

Main Results:

• Simulations indicated that low agent population densities and high mobility restrictions can significantly inhibit rapid pandemic spread.
• The framework provides a method for evaluating the effectiveness of interventions in urban settings.
• The hybrid neural network is designed for predicting the spatiotemporal probability of infection spread.

Conclusions:

• The proposed framework offers a novel approach to understanding and mitigating viral pandemic impacts in urban areas.
• Findings suggest that urban planning strategies focusing on population density and mobility control are crucial for pandemic resilience.
• This research aids urban planners and policymakers in designing sustainable and healthy post-COVID smart cities.