Camera-based occupant behavior trajectory capture and on-site assessment of spatio-temporal transmission risk of airborne virus in a university office building

Aerosol transmission is a route of contagion. The engineering field predicts risk in indoor environment by simulating the airborne transmission of the virus. However, occupant behavior is one of the main causes for the gap between predicted and actual airborne virus transmission risk. To improve the accuracy of transmission risk assessment, this study presents a framework for real-time assessment of infection risk based on the behavioral trajectory of the occupants captured by cameras. The study focused on analyzing the impact of short-range exposure and long-range exposure of occupants on the infection risk through fine-grained trajectories. On-site measurements were conducted in a university office building to obtain the air change rates (ACH) and occupant contact time respectively. A numerical model for quantifying the exposure risk of the two contact routes was also developed. The results demonstrated that the individual movement is diverse. The infection risk varies with the occupied space, behavioral patterns, and exposure time. The main passages and seats in the room, corridor corners, and room doorways are areas of the high frequency of short-range contact. The fraction of long-range exposure is 57.2%, indicating long-range exposure is the dominant route for the limited scenarios and population in this study. The cumulative risk of infection was 0.021 for teachers; 0.029 for employees; and 0.045 for students, with assumptions of initial infection rates of 5%. It can provide real-time warning for occupants through on-site monitoring and contribute to the transmission control and source tracing of airborne viruses.