An artificial intelligence (AI) model has been created by researchers at the Central Building Research Institute (CBRI) in Roorkee to forecast the "transmission probability" of COVID-19 in a closed area of a building. The model makes use of an electronic device to measure a room′s temperature, humidity, and carbon dioxide content. These and other input factors are used to display the likelihood that the COVID-19 virus will be present in a building′s office, classroom, or any other enclosed environment.

The findings of the study, titled "Transmission Probability of SARS-Cov-2 in Office Environments Using Artificial Neural Networks," were published in IEEE Access. Mr. Anuj Kumar, the principal investigator of the research, stated that the study is one of a kind.

According to the study, the R-value, or "anticipated number of new infections that originate in each event occurring over a complete duration in any space," was predicted using 11 input parameters. These parameters included CO₂ concentration (CO₂), area per person (AP), volume per person (VP), area per person (AO), number of inhabitants (O), indoor temperature (T_In), indoor relative humidity (RH_In), outside temperature (T_Out), and outdoor humidity (RH_Out).

Mr. Kumar added that some of these parameters and characteristics were estimated manually. The data was fed into two models: an artificial neural network (ANN), which is a computer-based mathematical modeling technique, and a curve-fitting model, which is a mathematical analysis tool.

The data showed that the correlation coefficients were 0.9557 for curve fitting and 0.9992 for ANN. He explained that the likelihood of the virus spreading increases as the value of these coefficients decreases.