Tangible GIS for Realtime Interactive Landscape Modeling

1
Tangible GIS for Real-time Interactive Landscape
Modeling

• Helena Mitasova,
• Marine, Earth and Atmospheric Sciences,
• North Carolina State University
• Lubos Mitas, Department of Physics, NCSU
• Russell S. Harmon, Army Research Office
• Carlo Ratti, SENSEableCity Laboratory, MIT

2
Topographic change
Elevation surface has been traditionally considere
d static for short time scales (days-years) Short
term topographic change Natural forces storms
in coastal areas, floods, landslides, gully and
stream bank erosion Anthropogenic construction,
agriculture, military New technologies are being
developed to monitor terrain change and
incorporate it into modeling and decision making
3
Anthropogenic terrain change
Construction reduced slopes, impermeable areas
added
Designed in CAD, represented by overlayed set of
contours difficult to work with
4
Preserve water flow pattern
discharge m3/s
School built, wetland and checkdams added for
stormwater control
GRASS GIS
Open Source Geospatial
Foundation
5
Maintain runoff

current 49 forest
construction 24 forest
discharge m3/s
GRASS GIS
Open Source Geospatial
Foundation
6
Background
New technologies are combining easy to
interpret 3d physical models of landscape with
geospatial data to facilitate communication and
collaboration
7
Current commercial options
GIS2MAP3D
XenoVision Mark III TerrainTableTM includes
TouchTable capabilities
8

Dynamic physical 3D models
Combine easy to interpret 3d physical models of
landscape with geospatial data to facilitate
communication and collaboration http//www.xenotr
an.com/ xenovision_clips.html
XenoVision Mark III Northrop Grunman
TerrainTableTM includes TouchTable capabilities
9
Motivation
In one-way systems re-designing terrain or its
features is complex, mouse and GUI
task MouseGUI we need to make a connection
between our hand and the image on the screen
10

New technologies Touch tables
IEEECGA Sept/Oct 2006 Interacting with
digital tabletops Interactive access
to geospatial information 2D representation Face
-to-face collaboration Limited design
capabilities
11
Illuminated Clay 3D interaction

developed by MIT Media Lab and SENSEable City lab
Combines virtual representation with solid model
and 2D images, re-computes and displays terrain
parameters in near-real-time
Traditional GIS design with mouseGUI need to
make a connection between hand and the image on
the screen Illuminated Clay hand and eye works
with the same object (physical model) freeing the
brain for more creative thinking
GRASS GIS
Open Source Geospatial
Foundation
12
Illuminated Clay 3D interaction
developed by MIT Media Lab and SENSEable City lab
Combines virtual representation with solid model
and 2D images, re-computes and displays terrain
parameters in near-real-time
Traditional GIS design with mouseGUI need to
make a connection between hand and the image on
the screen Illuminated Clay hand and eye works
with the same object (physical model) freeing the
brain for more creative thinking
GRASS GIS
Open Source Geospatial
Foundation
13
Two-way coupling with GIS
asynchronous
projector
color attributes
GIS/IC
3D scanner
project with delay
surface elevation
compute DEM and its parameters
scan and stop
modify by hand
synchronous- real-time response
projector
color attributes
GIS/IC
3D scanner
continuous loop
surface elevation
compute DEM and parameters instantly
scan continuously
modify by hand
14
Test study

explore how TanGIS can be used to solve
real-world problems common at communities and
installations investigate what new development
is needed to make the practical applications
feasible
20 ha area at NCSU exp. farms sediment and flood
control
GRASS GIS
Open Source Geospatial
Foundation
15
Analysis using a physical model

Data
Flow computed by Illuminated Clay in
synchronous mode
Aspect Slope
Physical Model
Shadow Elevation
Scanned phys. model
Slope and flow computed in GRASS in
asynchronous mode
GRASS GIS
Open Source Geospatial
Foundation
16
Design with TanGIS

adding a checkdam, creating a depression, while
watching the flow and slope to change
simulated overland flow depth for modified surface
GRASS GIS
Open Source Geospatial
Foundation
17
Design with TanGIS

adding a checkdam, creating a depression, while
watching the flow and slope to change
simulated overland flow depth for modified surface
GRASS GIS
Open Source Geospatial
Foundation
18
Building TanGIS at VISSTA lab

VIVID 910 laser scanner 1 scan/
0.3sec real-time interaction higher accuracy than
needed IR sensors cheaper, smaller need to be
tested
Multipurpose facility at VISSTA Lab at ECE NCSU
Prof. Hamid Karim
GRASS GIS
Open Source Geospatial
Foundation
19
Real-world and model DEMs

photogr.-based 2m DEM 1993
scanned model-based 1mm (2m) DEMs with various
modifications
lidar-based 2m DEM 2001
GRASS GIS
Open Source Geospatial
Foundation
20
Exploring runoff with TanGIS

Simulating flow over modified surface testing
algorithms exploring impacts
50cm
braided flow
Water depth 60cm
30cm
Smoothed real-world data
modified models
21
Future TanGIS

Physical model
Computer
color attributes
projector
GIS Real-time data from Terrestrial sensors and
satellites
3D scanner
surface elevation
3D shaper
Desktop or large collaborative systems
GRASS GIS
Open Source Geospatial
Foundation
22
Acknowledgment

This project has been supported by US Army
Research Office, NC WRRI and North Carolina
Sediment Control Commission We also thank Bill
Goran, USACERL for valuable discussions and ideas
on geospatial research for sustainable development
orolin.blog.sme.sk
23
Geospatial Design

Multidisciplinary team stakeholders and experts
- holds charrette intensive workshop - to set
goals, assess and modify the proposed project to
achieve high sustainability expressed by LEED
rating.
Traditional approach maps and slides Limited -
log from discussion - access to info - feedback
on impact of proposed modification Advantage fac
e to face,10 people
GRASS GIS
Open Source Geospatial
Foundation
24
Future needs for technology development -
innovative materials and methods for creating
computer and manually controlled, physical
models - materials and image processing methods
for real-time interaction with surface properties
25
Future needs for basic and applied research -
at what scales and for what type of tasks
are tangible physical models applicable, do we
need zoomable physical models (already
feasible with pin-based systems) - what is
the optimal approach for integration of GIS and
physical model data (accuracy, automated adjustmen
t and/or optimization of design, ...) - what
algorithms are needed for effective, real-time
coupling with dynamic models and real-time data