Democratizing Pedestrian Dynamics in Infection Modeling Applications
摘要
Pedestrian dynamicsPedestrian dynamics simulates the movement of individuals in a crowd. It has powerful societal applications, such as in steering public health policy and the design of the built environment. It can suggest fine-tuned interventions that mitigate infection risk without disrupting human activities. Broad use of pedestrian dynamicsPedestrian dynamics in such applications is hindered by the need for contributions from domains where experts lack substantial computing background. We discuss our solution through cyberinfrastructure developments, which ameliorates this problem by providing (i) a domain-specific language (DSL)Domain Specific Language (DSL) in which experts could provide input that gets translated to C++ code, (ii) a modular framework in which researchers could contribute models, (iii) a recommender systemRecommender system to enable users to determine suitable models, and (iv) a dynamically load balanced parallel parameter sweep engine to examine “what-if” scenarios. We then show results using pedestrian dynamicsPedestrian dynamics, which suggest effective public health interventions for air travel that can substantially reduce infection risk without reducing passenger traffic. The significance of this chapter lies in describing a solution to enable interdisciplinary contributions to complex pedestrian dynamicsPedestrian dynamics applications in public health applications.