Sensors and Crowd - Steve Liang, GeoCENS Project

Sensors Crowd

prepared for Cybera Summit 2010

Dr. Steve LiangUniversity of Calgary

and

Dr. Steve LiangUniversity of Calgary

!"#$%&$

– Dr. Steve Liang – http://sensorweb.geomatics.ucalgary.ca –

Time

log

(peo

ple

per

com

pute

r)

(Polastre, 2004)

Bell’s Law of Computing Classes

A new computing class every 10 years

Bell’s Law of Computing Classes

!"#$%&'()*+%,-.(*)/(0%--%#&1(#2(3/--(+4#&/1(2#)(

+/)1#&*-(3#$+56&'(*&,(74/(7/&1(#2(0%--%#&1(#2(

8%)/-/11(1/&1#)(&/71(7#(5&8%)/(*&,(%&7/)3#&&/37(

/9/).74%&':(;&(<=><?(*($*&(3#5-,(8*-@(%&1%,/(*(

3#$+57/)(*&,(0.(AB<B(:::(3#$+57/)1(*)/(0/'%&&%&'(

7#(C8*-@D(%&1%,/(#2(51E

'"(()*+,)*-../,*'"((0$*123*4%&*56"*'7&56*2#8*9"256*%4*:%;<=5"&*:(2$$"$)

"#$$5&%3*6#&1(#2(74/(F"G)*>2#=2&?*-../)*@%(=;"*ABCBD)*<<,*/EFGH

Bell, 2008

The Macroscope Vision

• The unprecedented amount of sensors will allow us to

• observe the world at very high spatial resolution

• perform these observations continuously

• collect them in digital form

• They will all be connected to the Internet

Burrell, J., Brooke, T. & Beckwith, R. Vineyard computing: sensor networks in agricultural production. Pervasive Computing, IEEE 3, 38-45 (2004). !

Vineyard Computing

– Dr. Steve Liang – http://sensorweb.geomatics.ucalgary.ca –

• Sensor/Data are scientists’ precious...

• They hide their precious...

• They don’t share their precious...

• But...their precious might be able to help others to save the world....

• And...even if they want to share and collaborate, there is no easy way...

Links between sensors are missing....

– Dr. Steve Liang – http://sensorweb.geomatics.ucalgary.ca –

Unleash the Power of Sensors

Innovation in Assembly

A Web of Sensors

Query the physical world from anywhere at anytime….

Sensor Web Vision

:&%38

www.OpenStreetMap.org

One of the most complete and most updated map databases in the world!!

2009-10-16

Quake Catcher Network

qcn.stanford.edu

classic.globe.gov

viz.globe.gov

Crowd may not the solution for every

problems...

But for some problems, crowd might be the best solution.

New Zealand Earthquake on Sep 3rd 2010

Crowd Characteristics

• Time Critical

• Rich Context

• Unlimited Growth

• Pareto Distribution

• Long Tail

Sensor Web Long Tail

Sensor Web Head

Crowd can do great work, if we can provide

them...

1. How to use low quality data to generate quality results

original GPS traces clarified GPS traces routable map

Cao, L. & Krumm, J. From GPS traces to a routable road map. Proceedings of the 17th ACM SIGSPATIAL International

Conference on Advances in Geographic Information Systems 3-12 (2009)

2. Search engine

• make it possible to find available data

3. Standards

• make it possible to access

4. Collaboration platform

• social networking

• able to upload/edit

Give the right tool to the crowd...

– Dr. Steve Liang – http://sensorweb.geomatics.ucalgary.ca –

alpha

alpha

Search, Auto Complete

and Suggestion

– Dr. Steve Liang – http://sensorweb.geomatics.ucalgary.ca –

Online Social Networks

friends

– Dr. Steve Liang – http://sensorweb.geomatics.ucalgary.ca –

Social Bookmarking

I"J"()*I,K*172#L)*!,K*)*M!NOFPOQ*OR5"#87#L*4%(S$%#%;7"$*5%*56"*!"#$%&*N"T)M*"#--*0#)*69/(H/34&#-#'%/1(*&,(I.17/$1(

J"HIK?(AB<B(;&7/)&*6#&*-(I.$+#1%5$(#&*)*@%(,)*#%,)*<<,UHGFUAE)*BVF-B*W2?*-.B.

– Dr. Steve Liang – http://sensorweb.geomatics.ucalgary.ca –

Recommendation cont.

Find people to find information

Find information to find people

(think eHarmony...)

Ping as an example...

GeoCENS Vision

Sensors

Workshop

- September 23rd 2010

- Great speakers:

• NASA

• OGC

• Microsoft Research

- Banff Centre

Dr. Steve Liang@geosensorweb