Multi-sensory integration for a digital earth nervous system

MULTI-SENSORY INTEGRATION FOR A DIGITAL
EARTH NERVOUS SYSTEM
FROM FRAMEWORK TO IMPLEMENTATION
Frank O. Ostermann (UT) & Sven Schade (JRC)
AGILE 2014 Conference, 05.06.2014

 Objectives
 Digital Earth Nervous System
 Multi-Sensory Integration
 From past work to design and implementation suggestions
 Future research and next steps
05.06.2014F.O.Ostermann & S. Schade - AGILE 2014 Conference 2
MULTI-SENSORY INTEGRATION FOR A DIGITAL EARTH
NERVOUS SYSTEM

(i) Stimulate and enrich the debate on interoperability for geospatial data
(ii) Increase understanding of the various interactions between geospatial
data collection, transformation, processing and usage on a global scale
(iii) Show potential future research foci
OBJECTIVES

 Overview
NERVOUS SYSTEM

 Real-time data input
through ubiquitous sensors
 Multi-layered, inter-
operable data sets
 Linked and open data
 Initiatives like GEOSS,
UNEP Live, …
THE BIG PICTURE: DIGITAL EARTH

CITIZENS AS SENSORS
+ = !
Why not treat information from the citizens
as another type of sensor data?

WHAT DOES THE CROWD DO? (GEOSPATIALLY, THAT IS)

WHAT’S HAPPENING NOW

 Overview
NERVOUS SYSTEM

Multi-Sensory Integration
 Combines the information from different sensory systems
 Results in coherent representation of the environment
 Is prerequisite for adaptive behavior and response to the environment
 Decreases sensory uncertainty and reaction times
Characteristics
 Mutual feedback between sensory systems
 Spatial proximity, temporal proximity, and inverse effectiveness
NEUROSCIENCE PERSPECTIVE

 Sensor data or information fusion
 Focus on
 Low-level abstracted sensor data
 Data fusion from several but similar sensors
 Different but related sensors in close spatial proximity (e.g. robotics)
 Less activity on
 Integration of heterogeneous sensors covering irregular areas
(hard/soft data integration from disparate sensors)
ENGINEERING PERSPECTIVE

 Two major principles from
neuroscience and cognitive
science align with core GI-
principles: what is near in space
and time is related
 Inverse effectiveness hints at
why outliers might be important
 Engineering provides methods
and algorithms
SO WHAT?

 Overview
NERVOUS SYSTEM

FRENCH FOREST FIRE SOCIAL MEDIA - GEOCONAVI

GEO-SOCIAL MEDIA AND CRISIS MANAGEMENT
New VGI
Relevant & Credible
VGI
VGI-Clusters
Fully Assessed
VGI & VGI-Clusters
Retrieval
(Web 2.0 APIs)
Assessment &
Enrichment
(Classification & spatial queries )
Clustering
(Spatio-temporal statistics)
Dissemination & Alerting
(WFS, WMS, SES, SMS, …)
What do
others
say?
What do
we
know?

DENS-MSI

• Cue combination
• Causal inference
• Assign likelihoods based on prior knowledge
• Challenges
• Choose relevant data sets (information retrieval)
• Train semi-autonomous system
• Crowd-sourced supervised machine learning
of a stratified sample of DENS perceptions
• Solution?
DEALING WITH CONFLICTING INFORMATION

 Overview
NERVOUS SYSTEM

LOW-COST IN-SITU AND MOBILE SENSORS
Publiclaboratory.com
Mikrokopter.de
Libelium
Waspmote

 Programming: Python
 Desktop Analysis: R, NLTK, Weka, SatScan, Postgis and Qgis
 Cloud Analysis: GeoCloud, GisCloud, CartoDB, AWS, Google Cloud,
Google Maps Engine, ArcGIS Online, GeoGit, HERE
 Communities and Projects: OSM, Zoomanitarians, UAViators, AIDR,
HDX, Crisismappers, Dolly, Pheme
HOW-TO AND TOOLS

NEXT STEPS
 Too specific approaches had little success in increasing interoperability
 Risk of failure for over-generic approaches
 Stepwise and incremental development methodology
 Starting from well examined cases, such as the forest fire or flood
examples for the initial set-up of a possible solution
 Moving into new areas such as urban environments
 Move to cloud processing

NERVOUS SYSTEM
Thank you very much for your attention!
Loooking forward to your questions!
f.o.ostermann@utwente.nl
sven.schade@jrc.ec.europa.eu

SOME PUBLICATIONS
 Craglia, M., Ostermann, F., & Spinsanti, L. (2012). Digital Earth from vision to practice: making
sense of citizen-generated content. International Journal of Digital Earth, 5(5), 398–416.
 Ostermann, F., & Spinsanti, L. (2012). Context Analysis of Volunteered Geographic Information
from Social Media Networks to Support Disaster Management: A Case Study On Forest Fires.
International Journal of Information Systems for Crisis Response and Management, 4(4), 16–37.
 Spinsanti, L., & Ostermann, F. (2013). Automated geographic context analysis for volunteered
information. Applied Geography, 43(9), 36–44.

Multi-sensory integration for a digital earth nervous system

Recommended

Recommended

More Related Content

Viewers also liked

Viewers also liked (19)

Similar to Multi-sensory integration for a digital earth nervous system

Similar to Multi-sensory integration for a digital earth nervous system (20)

Recently uploaded

Recently uploaded (20)

Multi-sensory integration for a digital earth nervous system