POPSIM: A Platform Targeting the Modeling and Simulation of Human Populations in Urban Environments

In this paper we describe a simulation platform, POPSIM, targeting the modeling and simulation of human populations in urban environments. POPSIM employs an agent-based modeling approach where individuals are intuitively represented by autonomous agents in the simulated population. Our efforts thus far have resulted in a prototype platform that embeds the following core functionality: a population synthesizer responsible for generating agent specific characteristics using demographic, geographic and relationship statistics; an extensible routine behavior model that use activity statistics to generate agent plans representing daily activities such as working, eating, leisure, sleeping; a growing set of critical city infrastructure models representing buildings, road-networks and cell-phone networks; and finally, a growing set of modular and reusable agent services such as activity planning and execution, field-of-view estimation, path-planning, opinion dynamics, etc.