ENS is revealing the previously unobservable in science applications

1
Embedded Networked Sensing Urban applications

• ENS is revealing the previously unobservable in
  science applications
• Multi-scale data and models to achieve context,
  and in network processing and mobility to achieve
  scalability (communication, energy, latency)
• Personal, social and urban sensing applications
  will exploit the millions of cell phone acoustic
  and image sensors
• Internet search, blog, and personal feeds, along
  with automated location tags, to achieve context,
  and in network processing for privacy and
  personal control

?
?
2
Core technology research

• From CENS research Relevance
• Multiscale sensing and actuation Ubiquity
• In-network processing Privacy
• Analysis and visualization Discovery
• To data-centric architecture Layer
• Observe, capture, forward Sensor
• Name, verify, tag with context Network
• Filter, search, store, disseminate Fabric
• Explore, task, re-present Application
• How to create opportunities for authoring and
  sharing at each layer?
• How to encourage responsibility from within the
  network?
• How to create re-usable web-based platforms?

3
Simple application examples

• From capturing moments to exploring rhythms
• Engaging the domain experts,
• Epiphanal vs. articulated data flows,
• Participatory design.
• Personal - Health monitoring
• Experiment with privileged but non-critical
  personal data.
• Social - Presence-based sharing
• Build local trust while retaining anonymity.
• Urban - Sound level mapping
• Participatory urban planning tool.

4
Common Sense Dana Cuff, Architecture and
Urban Planning Mark Hansen, Statistics and
DesignMedia Arts Jerry Kang, Law UCLA
5
Common Sense
Participatory Sensing Data Commons Building Places
6
CURRENT experiments with geoweb focus
on --lifestyle --individuals --information --int
ernet of things
Common Sensing
NEXT generation experiments with geo-sensing
should focus on --public sphere --groups of
individuals --response
7

• Participatory Sensing
• Top-down efforts to mobilize networks of embedded
  citizen-sensors
• Local expertise to help address global
• scientific questions
• Human identification and evaluation of phenomena
• Centralized data repositories allow citizen-
• sensors to share information and identify
  their
• contributions
• The connection to the physical world adds a
  dimension beyond the use of the web as
  communication channel physical phenomena, and
  their descriptions in data, become anchors for
  activity and collaboration
• Your participation, however is limited by the
  kinds
• of data you can share, and the interfaces used
  to
• organize or access the data the questions are
  
• determined for you

8

• Participatory Sensing
• Bottom-up mobilizations
• We can think of blogs and vlogs as a kind of
• local reporting from citizen-sensors
• From the sights of global conflicts and natural
• disasters, we have read reports by bloggers
• and other first-hand amateur journalists
• Google Maps mash-ups and other geography-
• based projects engage communities in sharing
• information participation requires only a
  mobile
• phone or PDA
• Sensing has even started to emerge in the
• media arts, projects that try to reveal
  something
• new about their neighborhoods automated,

9

• Infrastructure for sharing
• As we move from top-down science applications to
  citizen-initiated urban sensing, what precedents
  can we look to?
• How do we connect data like these to the
  internet? What
• lessons can we take from blogs and vlogs?
• Stretching things a bit, how do we enable
  sensor logs
• or slogs?
• To sense in the city is to reveal its
  inhabitants people may require some
  confidentiality protections before sharing
• What can the network do to aid in the
  anonymous or
• pseudonymous sharing of information? This is
• particularly important for acoustic sensors
  and imagers
• Are there configurations of hosting services
  and
• pub/sub models that will encourage
  participation?

10
(No Transcript)
11
Infrastructure of the Commons -implications for
emergent form -source of subsequent
problems -complications of retrofitting
Infrastructure of the Commons -choices are made
before the first bit is collected -organic DBMS
v. evidence-based practice -retrofitting
difficult, no one has the incentive
12

• Market Driven
• shopper profiling
• customized information
• target markets
• tagged products
• Citizen Driven
• customized information on demand
• spatially embedded
• political shopping

Contents of bag
Attire
Your mother looked at this card last week
75 of shoppers like you bought this card
Last Mothers Day you bought this card
13

• Design Experimentation in
• the Public Sphere
• responsive environments
• interactive urban interventions
• measuring and communicating
• connecting and participating

Meejin Yoon, White Noise/White Light, Athens
Lars Spuybroek, D-Tower, Netherlands
Diller Scofidio, Blur Building and Brain Coats,
Switzerlandd
14

• Distributed sense-making
• Who decides whats interesting in the data
  commons? Who asks the questions? How are results
  presented? Who ensures their trustworthiness?
  What action is taken as a result?
• If blogs spawned citizen-editors and
  journalists,
• what might we expect from easy access to data
• collection technologies, to publishing and
• collaboration?
• How can we contribute both through education as
• well as research, enabling more participatory
• users to make sense of extremely varied data
• types recorded under possibly unreliable
• circumstances
• Uncovering chance relations (data mining
• restored to its original usage?), reasoning in
  the
• face of uncertainties, recognizing poor quality
  data
• and data sources, drawing conclusions from

The clever data analyst need only expose himself
to what his data is willing (or even anxious) to
tell him
John Tukey
15
Embedded Networked Sensing Urban applications

• Core technology research
• Multiscale sensing and actuation,
• In-network processing,
• Analysis and visualization.
• City-scale applications
• Personal - Data shared with a privileged few,
• Social - Data circulated within existing social
  contracts,
• Urban - Data shared and viewed with varying
  anonymity.
• Design context and implications
• Inevitability - Participatory sensing becomes the
  norm.
• Adoption - Who frames the data commons?
• Provocation - What platforms spark communities?
• Engagement - How does this affect our civic lives?

16
Tentative class schedule

• Class Organization and Planning 4-3
• Review syllabus and reading list
• Go through slides (Google presentation subset)
• Assign reading and presentation tasks for Friday
  4-7 Find proposal and websearch urban sensing
  style apps (tagging, art, games, tourism, etc.)
• Discussion of acoustics and acoustic environments
  as a focus
• Discussion of reading sources (Ubicomp etc) so
  that we learn about the wheel and dont reinvent
• A proposed architecture for urban sensing 4-7
  (Andrew, August, Sasank)
• FIND proposal presentation overview Tech report
  to be posted

• Intro to urban sensing applications and platforms
  and infrastructure 4-10
• Note for those who can make it Special seminar
  10-11am Capkun 57-124 EIV. http//www2.imm.dtu.dk/
  sca/
• Urban apps paper(s) Tagging, starter list of web
  projects on http//www.lecs.cs.ucla.edu/urban-sens
  ing/index.php/Main_Page. Connection to and
  differences from science applications.
• Build on cellular infrastructure, handhelds, Read
  platform papers Nokia, 3G, handhelds (PAPERS
  TBD).
• Platform planning, coordination 4-12 (Note we
  will meet 3-4 in EE57-124 to listen to EE
  seminar, for those who can, and then only from
  4-5 in our regular room due to Passover)
• 3-4pm Special seminar, Lazos, Room 57-124 EIV
• 4-5 Discussion of platform for projects 770,
  maemo, python, google earth, acoustic tools, R,
  student groups sign up to become experts

17

• Wireless platforms and tools 4-17 (Martin, ??)
• Overview of wifi, gprs, mobile-apps, tcp over
  wireless,
• PAPERS TBD
• Platform tools
• http//nokia770.com
• http//www.teemuharju.net/2006/01/26/coding-for-no
  kia-770-using-python-part-1/
• http//www.python.org/
• http//diveintopython.org/toc/index.html
• http//www.maemo.org/
• http//semacode.org/
• Data in context (Georeferenced visualization,
  etc.) 4-19 (Stat students plus)
• Google earth, kml
• Tagging
• Other?
• PAPERS TBD

• Localization 4-26
• overview of gps, wifi based coarse localization,
  other such as bluetooth
• verified localization papers
• Acoustic environments 4-28 (makeup for missing
  4-24)
• Acoustic environment papers
• Will add image if students want

18

• Integrated topics 6 class sessions (May 1-19)
  will be student-group (approx 3 per group)
  designed in collaboration with instructor around
  a particularly interesting topic, project,
  platform, investigation.
• Topics to include acoustic environments survey
  and tools (2 sessions), privacy-preserving
  mechanisms, innovative tagging, interaction
  through georeferenced tools, authoring
  distributed applications, data acquisition
  techniques
• May 1,3 (Note May 4th is UCLA Urban Sensing
  Summit)
• May 8,10
• May 17, 19
• The last part of the class will focus on projects
  and interesting topics that come up
• May 24 Guest speaker, Peter Corke
• May 25 Makeup for May 22 TBD
• May 31 TBD
• June 5, 9 (makeup for june 7)
• Final project presentations

• Out of class work requirements
• prepare for the sessions that you are responsible
  for with your subgroup (includes iterating on
  locating material and coming up with coherent
  presentation)
• read materials for sessions you are not
  responsible for
• do weekly platform related implementation tasks
  in first part of quarter
• design and execute project with subgroup in 2nd
  half of quarter
• in class writing assignments per session (more or
  less)

19
Client
Client
Client
Client
Sensor
Subscriber
Mediator1
Mediator2
Registry
20
Sensor
Subscriber
Sensor
Subscriber
Sensor
Subscriber
Mediator1
Mediator2
Mediator1
Mediator1
Registry