geosimulation :: innovative geospatial simulation and analysis but innovative people

Home | Book | Research | Publications | Bio | Press | Geosimulation Labs
Dr. Paul M. Torrens, Department of Geography, University of Maryland, torrens at geosimulation dot com

Modeling crowd behavior

Publications are here | Project overview | Demonstrations | Eye candy | Support | Related groups
Project overview

This is an exciting time to be studying crowds. A confluence of developments of relevance is forming. Technologies are being introduced with unprecedented rates of development and uptake and with unforeseen influence on the ways crowds interact with the city. Data and the dataware that generates them are available in new forms and in new quantities. New perspectives on old views are reshaping the way we theorize. A new policy environment exists for consideration of downtown environments, and crowds as their lifeblood—old topics of downtown revitalization are being revisited, and new foci related to homeland security and defense are coming to the fore. All of this is taking place under the umbrella of new emerging trends and behaviors in a larger societal context. Nowhere is this more relevant than at the micro-scale, on the streets, in and around our downtowns, and among the crowds of people that populate and energize our environment. For cities, a new appreciation of urban geography is gathering steam, an urban geography of the micro-scale, where pedestrians swarm in social and anti-social networks; where innovative Information and Communications Technologies (ICT) are being deployed at street-level, digitally-enabling crowds through networked computing. Embedded in urban infrastructure and in the very products we consume, the same technology allows cities to think about—and process—the people that pulse through them.

 
The goals of this project are to build a reusable platform for modeling human behavior, action, and interaction in social and anti-social crowds, for the purposes of simulating a variety of behavioral, human, and urban geography scenarios. Methodologically, the tool-smithing for this project is advancing on the idea of Geographic Automata, which are used as the building-blocks for modeling crowds from the bottom-up. Modeling tools are being tightly-coupled to space-time Geographic Information Systems and social network analysis, for visualization puposes, but also for behavioral analytics. Substantively, simulations are being constructed around a variety of theory-driven scearios, with strong practical currency: human activity spaces, navigation and wayfinding in urban environments, complexity signatures in dynamic and adaptive socio-spatial systems, rioting and civil violence, hazards and emergency evacuation, crime and defensible space, retailing and business geographics, among others.
 

See our projects on immersive modeling also.

See my lecture on the topic here.

Demonstrations
 
Simulating crowd behavior (you will need the Adobe Flash Player plug-in for your browser to view this movie)
 
Eye candy
(Print-quality versions of these graphics are available upon request)

A crowd streams through a dense urban setting

A bottleneck forms in an urban canyon

The simulation run with low polygon actors

The simulation can be run from any vantage point

crowd evacuation
The simulation will run agnostically relative to fixed infrastructure. This illustration shows the model applied to an indoor concert hall evacuation scenario.

path-planning

The model treats high-level, medium-level, and low-level behavioral geography. Above, the determination of a space-time path between a trip source and sink is displayed, as an example of high-level behavioral geography. A variety of path-planning algorithms can be used.

wayfinding

Medium-level behavioral geography: above, wayfinding between fleetingly-resolved landmarks (lamp-post markers on street corners, in this case) is shown.

 

locomotion and steering

At a lower-level of behavioral geography, steering and locomotion are used by agents to articulate their movement through space and time, while seeking-out or avoiding collisions with fixed and mobile objects.

Probes in the simulation report data every 1/60secs to a space-time GIS for processing

Post-simulation analysis highlights the sources of gridlock in the crowd and the urban infrastructure

Support
Torrens, P.M. (2007-2012) “CAREER: Exploring the dynamics of individual pedestrian and crowd behavior in dense urban settings: a computational approach”. National Science Foundation (Faculty Early Career Development (CAREER); Geography & Regional Science/ Methodology, Measurement, and Statistics)
 Torrens, P.M. (2006) “Urban simulation at the micro-level”. Autodesk, Inc.
Related groups
GV2, Trinity College Dublin
Design Graphics Lab, North Carolina State (Benjamin Watson)
Digital Phoenix, Arizona State University

 

Projects >>

Dynamic physics for built infrastructure

moving agents through space and time

Moving agents through space and time

modeling riots

Modeling riots

Validating agent-based models

Machine-learning behavioral geography

Accelerating agent-based models

megacity models

Megacity futures

Immersive modeling

Space-time GIS and analysis

A toolkit for measuring sprawl

space-time GIS

Modeling time, space, and behavior

simulating crowd behavior

Simulating crowd behavior