We present a method for populating procedurally generated 3D city models with crowds of artificial agents. It is targeted towards the analysis, prediction and visualization of occupant behaviour in urban planning. We simulate and quantify correlations on the following aspects: functions of buildings, number of people and fluctuation in density. Potential practical applications are for example a) to determine bottlenecks in public transit, b) to identify possible problems for evacuation scenarios, c) to evaluate the demand for and the accessibility of amenities as well as d) the stress of pedestrians to evaluate quality of life indicator for a given urban region . The occupants’ location data – represented by the agents - and relevant semantic metadata are encoded inside a grammar-based city modelling system. This information is used for the context-dependent automatic placement of occupant locators during the procedural generation process of the urban 3D model. Most of the underlying parameters are interconnected with each other. For example, the number of resulting agents corresponds to the size, function and location of one specific building. Once a 3D city model has been generated, occupants are represented by agents using a) a commercial fuzzy logic system and b) pre-animated 3D avatars. The agents find their way through the city by moving towards points of interest to which they are attracted. Each individual agent draws specific paths while interacting with the urban environments and other agents. Every path describes a set of parameters, for example speed, space available and level of exhaustion. The ensuing visual diagrammatic representation shows the resulting agent paths in correlation with the virtual environment. This offers the opportunity to investigate parts of a city and optimise corresponding aspects with minimal interventions on the urban level. We show the application of this method to evaluate planning interventions in the urban fabric and monitor the correlating effects.
[1]
Daniel Thalmann,et al.
Handbook of Virtual Humans
,
2004
.
[2]
Antje Kunze,et al.
Using Shape Grammars for Master Planning
,
2008
.
[3]
Yiorgos Chrysanthou,et al.
Real-Time Rendering of Densely Populated Urban Environments
,
2000,
Rendering Techniques.
[4]
Remo Burkhard,et al.
ETH Value Lab: A framework for managing large-scale urban projects
,
2008
.
[5]
J. Gibson.
The Ecological Approach to Visual Perception
,
1979
.
[6]
Pascal Müller.
Procedural modeling of cities
,
2006,
SIGGRAPH Courses.
[7]
Luc Van Gool,et al.
Procedural modeling of buildings
,
2006,
SIGGRAPH 2006.
[8]
Daniel Thalmann,et al.
Hierarchical Model for Real Time Simulation of Virtual Human Crowds
,
2001,
IEEE Trans. Vis. Comput. Graph..
[9]
Wander Jager,et al.
Clustering and Fighting in Two-party Crowds: Simulating the Approach-avoidance Conflict
,
2001,
J. Artif. Soc. Soc. Simul..
[10]
Roger Diener,et al.
Switzerland - an Urban Portrait
,
2005
.
[11]
Alan Penn,et al.
Space syntax based agent simulation
,
2001
.