Paul Torrens, Modeling Crowd Behavior
by Kurt Cagle
Paul Torrens (Arizona State University)--Ambient crowds are the new distributed computing platform. Smart mobs are fashioning new architectures for social networking. Armed with cell phones and mobile gaming devices, they are the new business model for location-based services. Seditious crowds are creating havoc in urban theaters of war and at global economic forums. Crowds of shoppers, endowed with smart chip credit cards and RFID tagged merchandise are trailed by long-lasting data shadows that follow them ubiquitously.
Embedded in urban infrastructure and in the very products we consume, new technologies are emerging to enable cities to think about—and process—the people that pulse through them, with a burgeoning code-space being developed to capture the actions and interactions of individuals within large dynamic crowds. This presentation will focus on our recent research work in developing models of crowd behavior and their application to theory-building and scenario evaluation in the contexts just described.
We have developed a reusable modeling platform for constructing large simulations of individual and collective behavior in dense urban environments. The simulations are developed with individual agents, equipped with geospatial AI that allows them to perceive and react to their evolving surroundings with an incredible level of behavioral realism. These agents are also capable of social and antisocial interactions. The simulation architecture is coupled to Geographic Information Systems, allowing for a suite of geospatial analytics and data-mining to be performed, across a wide array of scenarios. Moreover, the models have been developed as realistic 4D immersive environments with unprecedented levels of graphical realism.
From O'Reilly Where 2.0, San Jose, CA, Tuesday, May 29th, 2007.