Improving Pedestrian Flow through Modeling and Space Syntax

Report
A Case Study of the Wings Over Pittsburgh Air Show
Benjamin Shelton - MGIS Candidate
Advised by Dr. Alexander Klippel
Motivation
Wings Over Pittsburgh Air Show
Photo Courtesy of http://www.pittsburgh.afrc.af.mil/
Motivation
Wings Over Pittsburgh Air Show
Photo Courtesy of http://www.pittsburgh.afrc.af.mil/
Motivation

Improve the placement of static
displays, tents, and vendors.
 Improve accessibility for general public.
 Identify major thoroughfares to ease
flow of pedestrians.
 Improve egress routes for
disaster/emergency preparedness.
Research Question
How can geospatial modeling and
analytical methods to represent
pedestrian movement be incorporated
to improve the flow of pedestrians and
accessibility of a large public event?
-Flow measured through use of Agent Based Modeling scenarios.
-Accessibility measured in terms of Space Syntax parameters.
Methodology
1.
2.
3.
Determine best methods for visualizing
and analyzing pedestrian movement
through a space.
Determine how to best incorporate into
GIS.
Assessment in real world scenario
using past Air Show data.
1. Methods for visualizing & analyzing pedestrian flow

Space Syntax
 Science-based approach to urban planning, design, and
architecture.
 Characterizes spaces to analyze movement, interactions, land uses,
and many other aspects.
 Axial Lines:
 Used to represent flow of movement along lines of visibility from point
to point through a space.
 Used to measure spatial connectivity and integration.
1. Methods for visualizing & analyzing pedestrian flow

Space Syntax Example
[2011 Air Show Layout]
1. Methods for visualizing & analyzing pedestrian flow

Space Syntax Example
[2011 Integration]
1. Methods for visualizing & analyzing pedestrian flow

Space Syntax Example
[2009 Integration]
1. Methods for visualizing & analyzing pedestrian flow

Space Syntax Example
[2009 Connectivity]
1. Methods for visualizing & analyzing pedestrian flow

Space Syntax Example
[2011 Alternate Layout Integration]
ADJUSTED
ORIGINAL
1. Methods for visualizing & analyzing pedestrian flow


Agent-Based Model (ABM)
 Agent: Autonomous, goal-directed data points that are
interactive, both with the environment and with other agents.
Pros:
 More realistic than previous density or group-level approaches for pedestrian
behavior .
 Modeling at the individual level provides insight into complex local interactions
that influence the overall movement pattern.
 Flexible to fit many scenarios and environments.

Cons:
 Requires high computer processing to handle a large number of agents.
 Difficult to encompass all the complexities of human behavior into an agent.
2. Incorporating into GIS

Agent Based Models:
 Hurdles to implementing ABM completely within GIS:
 Difficulty of GIS in handling temporal data.
 Inability of GIS to handle the execution of a model of agent behaviors
for individual data points.
 GIS-centric vs. ABM-centric
 GIS-centric: External tools and scripts used within GIS software to
develop an Agent-Based Model.
 ABM-centric: GIS data used within external Agent-Based Modeling
software.
**PedestriansV1 ABM example downloaded
from NetLogo User Community Models Website.
Originally created and submitted by Singhathip
Mickael in 2008.
Possible ABM Software Solutions:

Swarm
 Open source
 Objective C or Python

Repast (“Recursive Porous Agent Simulation Toolkit”)
 Open source
 Java or Python
 Agent Analyst toolbox for ArcGIS – Latest release is for 9.2 (GIS-centric example)

MASON (“Multi-agent Simulator of Neighborhoods”)
 Open Source
 Java

NetLogo
 Open Source
 Has GIS extension that provides support for vector and raster data integration.
Possible Space Syntax Software Solutions:

UCL Depthmap
 Standalone product.
 Able to import geometry of environment and export space syntax results into
MapInfo or .DXF file format .

Confeego
 Integrated with MapInfo.

Axwoman
 Works as an extension within ArcGIS (Originally for ArcView 3, but latest release
works within ArcGIS10).

Others include Syntax2D, Place Syntax Tool, Segmen

Currently, External software provide more robust analysis and documentation than
any GIS-based Space Syntax solutions.
Process:
 Agent-Based Model:
 Used to model scenarios of pedestrian movement through the air show space
 Space Syntax:
 Used to characterize the layout of the air show space and explain the
movement shown in the Agent-Based Model.
ArcGIS
Export Layout
2011 Air Show
GIS Layout
Export Layout
NetLogo
UCL
Depthmap
ABM
Simulation
Visibility Graph
Analysis
•Pedestrian Flow
•Simulated gate counts
•Integration Values
•Connectivity
Process:
Perform Space
Syntax Analysis
-Integration Values
-Connectivity
2011
Perform ABM
Simulation
-Integration Values
Change in
Integration/Connectivity
& Model
-Connectivity
2009
Perform ABM
Simulation
Perform Space
Syntax Analysis
-Pedestrian Flow Model
-Integration Values
Change in
Integration/Connectivity
& Model
-Connectivity
2008
Perform ABM
Simulation
Perform Space
Syntax Analysis
-Pedestrian Flow Model
-Integration Values
-Connectivity
ALT
Perform ABM
Simulation
-Pedestrian Flow Model
Change in
Integration/Connectivity
& Model
Find Correlations
Perform Space
Syntax Analysis
-Pedestrian Flow Model
Conference Presentation / Project Timeline
7th International Conference on Geographic Information Science
(GIScience 2012)
•
•
•
September 18-21, 2012
Deadline for Extended Abstract Submittal :
•
May 4, 2012
Questions?

похожие документы