Predicting COVID-19 Infection Risk in Multi-Use Facilities Using AI

Nuclear Research Institute Develops 'COVID-19 Transmission Risk Prediction Simulation for Multi-use Facilities'

Results of the Korea Atomic Energy Research Institute's simulation of COVID-19 infection at the Guro call center using artificial intelligence. Healthy individuals (blue human figures), infected individuals (red), and individuals in the incubation period (yellow). (Photo by Korea Atomic Energy Research Institute)

[Asia Economy Reporter Moon Chaeseok] As small-scale infections of the novel coronavirus disease (COVID-19) occurring in multi-use facilities such as aerobics studios, academies, PC rooms, and saunas lead to large-scale outbreaks, a technology capable of predicting the COVID-19 transmission risk in multi-use facilities has been developed.

The Korea Atomic Energy Research Institute announced on the 1st that it has developed a 'Multi-use Facility COVID-19 Transmission Risk Prediction Simulation' technology in collaboration with private companies specializing in artificial intelligence and big data.

According to the Atomic Energy Research Institute, while there are mathematical models simulating COVID-19 transmission patterns at the national level, there is no technology to predict the infection risk of the facilities themselves that are used by many people.

The research team modeled the space of multi-use facilities and the movement paths of individual users using Unity. Unity is software mainly used to create content such as 3-dimensional (3D) animation, architectural visualization, and virtual reality (VR).

Based on this, the overall risk for users and the facility was calculated according to the distance between people through a transmission probability model. The transmission probability model refers to a probabilistic model that predicts infection based on each individual's symptoms and the distance between individuals.

When the research team applied the newly developed simulation to the previous Guro Call Center infection case, results similar to the epidemiological investigation paper by the Korea Disease Control and Prevention Agency were derived.

It also proved that when the density of users was the same, facilities with more overlapping movement paths had higher infection risks.

Yoo Yonggyun, head of the Intelligent Computing Research Laboratory at the Atomic Energy Research Institute, who led this study, said, "Using simulation technology, it is possible to predict the risk considering factors such as the facility's population density and mask-wearing status to find the optimal quarantine policy."

Yoo added, "This study was developed based on simplified rules and is not a study involving epidemiologists or medical professionals. It is necessary for various experts to participate to review and enhance the accuracy of the simulation model."

This content was produced with the assistance of AI translation services.