Numerical model for prediction of indoor COVID-19 infection risk based on sensor data

In addition to infection with SARS-CoV-2 via direct droplet transmission or contact with contaminated surfaces, infection via aerosol transport is a predominant pathway in indoor environments. The developed numerical model evaluates the risk of a COVID-19 infection in a particular room based on measurements of temperature, humidity, CO2 and particle concentration, the number of people and instances of speech, coughs and sneezing using a dedicated low-cost sensor system. The model can dynamically provide the predicted risk of infection to the building management system or people in the room. The effect of temperature, humidity and ventilation intensity on the infection risk is shown. Coughing and especially sneezing greatly increase the probability of infection in the room; therefore distinguishing these events is crucial for the applied measurement system.