Indoor environments can pose a substantial respiratory infection risk, especially in densely populated areas. It is therefore vital to evaluate and predict infection risks for mobile individuals engaging in indoor activities. However, previous studies have primarily focused on assessing viral transmission probabilities between relatively stationary individuals, disregarding the dynamic features of pedestrian trajectories. To address this issue, we propose an approach that uses precise trajectory data and digital twin technology to evaluate and forecast indoor infection risks for mobile individuals. Our approach involves the utilization of high-precision trajectory data to develop a time-varying virus density field map (VDFM), forming the basis for infection risk assessment. Additionally, we present a deep-learning model that employs the transformer method to predict future time-varying VDFMs and associated infection risks. Furthermore, we introduce digital twin to comprehend real-time interactions between the physical and digital realms within the structure. To validate our approach, we conducted a case study at the Wuhan International Exposition Center, serving as the control room. We performed a sensitivity analysis on various preventive measures, including mask-wearing, social distancing, and indoor infection control. The results highlighted the significant impact of these measures on individual infection risks associated with indoor activities. Our research contributes to the development of tailored prevention and control strategies, which play a critical role in mitigating the spread of respiratory infections. Thus, this study holds significant implications for technology-facilitated public health services.
Read full abstract