We are building a Chelsea Policy Simulator Tool that will allow decision makers to intuitively understand the effect of different COVID-19 policy intervention retrospectively in 2020. For example, the policy simulator will show the impact of delaying mask mandates or reducing overcrowding in the city.
Behind the scenes, the simulation tool consists of a series of agent-based, data-driven models of mobility, infection, testing and disease progression. It allows for the simulation of individual people as, every day, they commute from home to work and school, and visit points-of-interest such as grocery stores, and bars.
We use a series of prior and posterior predictive checks to calibrate the model against COVID-19 testing, case and death data collected by the state and against estimated cases we infer through wastewater sampling.
This effort integrates approaches in Bayesian statistics, infectious disease epidemiology and agents-based simulation. Our goal is to empower Chelsea city policy makers and residents for better interventions and advocacy.