Computational Model for Pedestrian Movement and Infectious Diseases Spread During Air Travel: A Molecular Dynamics-Like Numerical Approach

There is direct evidence of the transmission of fatal infectious pathogens in large human gatherings. Air transportation is no exception. The mixing of susceptible and infectious individuals in this high-density man-made environment involves pedestrian movement which is generally not taken into account in modeling studies of disease dynamics. This thesis addresses this problem through a multiscale model that combines pedestrian dynamics with stochastic infection spread models. This generic model is applicable to several directly transmitted diseases. Through this multiscale framework, the effectiveness of certain layout and strategies in suppressing the disease spread in highly crowded locations such as airplanes, airports and waiting queues is quantified. Inherent variability in human behavior leads to a larger parameter space. This large parameter space is addressed by using novel parallel algorithms for parameter sweep based on low discrepancy parameter sweep, compared to a default lattice-based sweep. This dissertation shows that certain pedestrian movement strategies may be adopted during an outbreak to reduce pedestrian-to-pedestrian contacts. For instance, two-section boarding leads to lower infections whereas all deplaning strategies have a similar effect. Winding queues configurations at security checkpoints or theme parks have a major effect on pedestrians’ interaction. A queue of two-zones with two inlets and outlets and vertically portioned short aisles is superior over the other assessed configurations in terms of reduced infection. In terms of parameter sweep of the large domain, a low discrepancy Halton sequence is used for uncertainty quantification. This method has proven to be efficient and less time consuming when applied to at least one model of the entire multidisciplinary model compared to the default lattice-based model