Sensor%20Networks%20in%20the%20Wild:%20Challenges%20and%20Opportunities%20for%20Semantic%20Web%20Technologies

1
Sensor Networks in the Wild
Challenges and Opportunities for Semantic Web
Technologies
David De Roure
2
Sensor Networks in the Wild Challenges and
Opportunities for Semantic Web Technologies

• David De Roure

3
Overview

• Daves Gently Holistic Detector Agency
• Ice and mud Glacsweb and Floodnet
• Human in the loop the planet and the people
• Instrumented citizens
• Challenges and opportunities
• Context, provenance
• Processing, automation
• Configuration, diagnostics
• Annotation, notification
• Assembling applications
• Method deluge, curation
• Discussion

4
Which one of these is a sensor network?
5
(No Transcript)
6
You havent done really sensor networks research
until...

• You have fallen in the river/mud/glacier/...
• You have felt the excitement of data arriving.
  And uncertainty when it stops.
• The data has had an impact.
• Methodologically, this research must be
  conducted in the wild.

7
But why is it worth falling in mud?

• Intelligent sensor networks in an environmental
  application address an important set of grand
  challenge Computer Science issues including

• Scale, scalable
• Autonomic behaviour versus control
• Persistent, heterogeneous, evolving
• Holistic approach including information systems
• Deployment challenge
• Some mobile devices

8
Grand Challenge
Scalable Ubiquitous Computing Systems

• We need a collection of principled techniques
  that apply to systems with many more nodes,
  smaller as well as larger ones, and a much larger
  range of capabilities per node, per cluster of
  nodes, all carrying out a much wider range and
  number of different tasks

• From the model/architectural viewpoint, a goal
  should be to provide a hierarchy of abstractions
  that allow us to specify and understand these
  systems at many levels of detail
• Grand Challenge Proposal 4(Jon Crowcroft)

9
Unsolved issues
Challenges

• Context awareness
• Trust, security and privacy
• Seamless communication
• Low powered devices
• Self configuration
• Information overload
• Information provenance
• Support tools non-existent
• Human factors
• Social Issues
• Business Models

• Convergence (consistency)
• To choose range of consistency and to reason and
  manipulate convergence goals
• Relevance
• Achieving global temporal and spatial locality
• Cooperative filtering
• Correctness
• Assurance and availability
• Authoritative source, provenance and
  meta-information

UKCRC Grand Challenge
10
FloodNet
8
9
7
10
2
12
1
3
11
4
6
5

• Yellow spots indicate the location of sensors

11
River Crouch study site
Standing at sensor 5 facing North East
overlooking Brandy Hole
12
Tidal channel at low and high tide
13
Solar powered project!
14
Spill unit
Distance between sections
Channel section
ISIS Flow model
15
FloodNet Architecture
The inner loop
subscribers
Simulated Nodes
users
Peer-to-peer computing
GIS
gateway
broker
GPRS
Flood prediction
Sensor Nodes
Live data informs predictions
grid
Predictions influence sampling and reporting
rates
Meteorological data
The outer loop
16
Impact of information
17
Towards a global environmental sensor network
18
www.myocean.eu.org
19
(No Transcript)
20
Chawton House and Landscape
Centre for the Study of Early English Women's
Writing

• School children using PDA and GPS
• Focus on different users tools for experience
  builders
• Accumulation of content and annotations
• Record and reuse

21
Developing software solutions to provide timely
video supported feedback to nursing students
after simulated practice events
Big Sister
22
Keep in mind 3 users...

• The climate change researcher
• The citizen
• with RFID tag, phone, MP3, GPS, ...
• The sensor network support guys

23

• As data is collected we need to record the
  context so that it may be properly interpreted
• Where
• When
• What
• What units
• How e.g. Device type, id, config, calibration
• What error e.g. how many satellites?
• SensorML

24
(No Transcript)
25

• When data is published we need to know its
  provenance so that we can understand quality and
  trustworthiness
• Helps choose between multiple sources
• Realtime, near-realtime, validation
• e.g. Open Street Map vs Ordnance Survey

26
Open Provenance Model
Communications of the ACM 51, 4 (Apr. 2008), 52-58
Scientific Discourse Relationships Ontology
Specification
27

• With more devices delivering more data more
  often we need to deal with a deluge of data
  through automation
• Data-intensive science
• Citizen Science
• Semantic Web facilitates automation... but thats
  another talk!

28
E. Science laboris

• Workflows are the new rock and roll
• Machinery for coordinating the execution of
  (scientific) services and linking together
  (scientific) resources
• The era of Service Oriented Applications
• Repetitive and mundane boring stuff made easier

29
(No Transcript)
30

• We need devices and networks to be configured
  automatically

Input 0
Data link
USB interface
31
(No Transcript)
32
(No Transcript)
33

• We need to be able to monitor devices and
  networks, and to handle diagnostics

34

• Data is linked, annotated and integrated

People and sensors annotate space. Share the
annotations!
35
Imagine these are all live feeds with archives
36

• Its about live data and notifications
• Pub-sub would benefit from ontologies
• RSS and Atom
• So would log file analysis

37
Automatic Blogging by Machines
38

• Mashups and workflows can be assembled rapidly
  with assistance in finding and integrating data
  and the services that process it
• Rapid application development is about end user
  programming
• Mashups are good now
• What happens when there are millions of services?

39
(No Transcript)
40

• Research question
• Does Semantic Web make it easier to assemble the
  applications?

41

• We share not just the data but the methods that
  are used to handle it
• Data deluge brings method deluge too
• Community sharing
• Associate methods with data

42
(No Transcript)
43
Paul Fisher
Workflow 16
QTL
Results
Logs
produces
Included in
Published in
Included in
Feeds into
produces
Included in
Included in
Metadata
Slides
Paper
produces
Published in
Common pathways
Results
Workflow 13
44
http//rdf.myexperiment.org/Aggregation/Pack/56
45
Curation of data, services and methods
Self by Service Providers
Experts
refine validate
refine validate
seed
seed
Workflows and Services
refine validate
refine validate
seed
seed
Social by User Community
Automated
46
This projects aims to shed light on patterns in
social dynamics and coordinated human activity.
We do so by developing and deploying an
experimental social interaction sensing platform.
This platform consists of portable sensing device
and software tools for aggregating, analyzing and
visualizing the resulting data.
www.tagora-project.eu
47
(No Transcript)
48
Discussion

• More sensors, more networks, more data and of
  different quality, and more overlapping as time
  goes on
• Many of the very real challenges in the future of
  sensor networks may benefit from Semantic Web
  technologies not just sensor data
• It is fundamentally about record and reuse and
  unanticipated reuse!
• Characterised by multiple sources of interlinked
  structured data
• Annotate the planet ?
• There will not be one giant triplestore the size
  of the planet the triplestore is the planet! But
  finding stuff matters.
• Good circumstances in the ecosystem bottom up
  and top-down forces, standards like OGC and W3C,
  some low-hanging fruit
• PS Come to the ISWC Semantic Sensor Network
  Workshop!

49

• Contact
• David De Roure
• dder_at_ecs.soton.ac.uk
• Slide Credits
• Kirk Martinez, Alex Rogers, Jeremy Frey, Carole
  Goble, Mark Weal, Chris Greenhalgh, Jon Blower,
  Oscar Corcho, Paul Fisher, Jacob Beal