1. Crowdsourcing Urban Data with Smartphones
Demetrios Zeinalipour-Yazti
Data Management Systems Laboratory
Department of Computer Science
University of Cyprus
Invited Talk at the Mining Urban Data (MUD) Workshop
(with EDBT/ICDT), Athens, Greece, March 28, 2014

2. Talk Objective
To review some primitive web crowdsourcing concepts and challenges.
To show how these challenges emerge and evolve in Urban Data Collection spaces.
To present some of our own developments related to: i) Location Data; ii) Data Collection Testbeds, and iii) Social Data and discuss particular challenges and future work.
Much of the discussion is work in progress (how we plan to apply the ideas in urban spaces).
IEEE MDM’13 Tutorial:
"Crowdsourcing for Mobile Data Management", G. Chatzimilioudis and D. Zeinalipour-Yazti, "Proceedings of the 14th International Conference on Mobile Data Management" (MDM '13), Milan Italy, Volume 2, Pages: 3-4, 2013.

3. Crowdsourcing Definitions
Crowdsourcing = Crowd + Outsourcing
Jeff Howe (2006). "The Rise of Crowdsourcing". Wired.
From our recent work:
Crowdsourcing refers to a distributed problem-solving model in which a crowd of undefined size is engaged in the task of solving a complex problem through an open call for monetary or ethical benefit.
“Crowdsourcing with Smartphones”, Georgios Chatzimiloudis, Andreas Konstantinidis, Christos Laoudias, Demetrios Zeinalipour-Yazti, IEEE Internet Computing, Special Issue: Sep/Oct Crowdsourcing, May IEEE Press, Volume 16, Pages: 36-44, 2012.

4. Web Crowdsourcing Platform Workers Requester Rewards Open Call (Task)
Solutions
Requester
(Crowdsourcer)
Rewards
Workers
(Solvers)
Platform

5. Web Crowdsourcing Microtasking Platform: Qualifications a) Reward
b) Redundancy: Each worker solves a Hit once (3-5 assignment per hit) to enable majority voting
a) Reward

6. Web Crowdsourcing: Incentives
Tangible (Monetary) Incentives
Cash, Credit or Gifts (MTurk, Kickstarter)
Unintended or as-a-by-product (reCaptchas)
Ethical Incentives
Socialize & Fun
Earn Prestige
Altruism
Learn something New
Usually a combination of several incentives

7. Web Crowdsourcing: Challenges
How to Recruit Contributors (randomly, marketplaces?) / What the Contributors Can Do (qualifications, tests)?
How to Combine their Contributions?
How to Manage Abuse?
How To Scale/Manage Complex/Larger Tasks?
Openness / Quality?
Disclosure Issues (Privacy related to Tasks, NDAs?)
Minimum Wages & Social Contributions?
Anhai Doan, Raghu Ramakrishnan, and Alon Y. Halevy Crowdsourcing systems on the World-Wide Web. Commun. ACM 54, 4 (April 2011),

8. Declarative Crowdsourcing
CrowdDB, Qurk, Deco, MoDaS, Crowdforge.
SELECT abstract
FROM talk
WHERE title = "CrowdDB";
Crowd Extensions
CrowdDB: Answering Queries with Crowdsourcing,M. J. Franklin, D. Kossmann ,T. Kraska, S. Ramesh, R. Xin, SIGMOD‘11 & VLDB'11Demo

9. Mobile Crowdsourcing
txtEagle (now JANA) founded by Nathan Eagle (PhD, MIT, 2005) a first-of-a-kind mobile CS system:
Requesters: can assign small tasks (translation, transcription and surveys) on their mobile phones.
Workers (today 3.48 Billion Workers in 102 countries!): : rewarded with airtime on their mobile subscriber accounts or MPESA (mobile money described next).
txteagle: Mobile Crowdsourcing, Internationalization, Design and Global Development, LNCS Volume 5623, pp , 2009.

10. Mobile Crowdsourcing Another app txtEagle SMS Bloodbank :
Idea: to report blood levels of local hospitals centrally by nurses.
Initially, in the absence of an incentive, the system was a complete failure.
In summer 2007, automatic airtime credit was incorporated to award nurses for their contribution => then a huge success!
Other txtEagle SMS applications:
Transcription mentioned previously (global market $18B in 2010)
Software Localization (60 local languages in Kenya, txtEagle generated a cookbook
Citizen Journalism, Sentiment Analysis, Surveys

11. The Smartphone Era April 2013: Beginning of Smartphone Era!
In April, 2013, for the first time in history the number of Worldwide Smartphone sales exceeded that of feature phones (according to IDC)
51.6% were Smartphones (216M units)
48.4% were Feature Phones (186M units)
The bulk of mobile phones are acquired in the developing world (e.g., China, India, Africa etc.)
Chinese manufactures (ZTE, Huawei) started building smartphones for the wide markets.
More Smartphones Were Shipped in Q Than Feature Phones, An Industry First According to IDC, 25 Apr 2013,

12. Crowdsourcing with Smartphones
A smartphone crowd is constantly moving and sensing providing large amounts of opportunistic data enabling new applications

13. Smartphone Crowdsourcing: Challenges
Challenges (Beyond Web Crowdsourcing)
Big Data
Velocity by sensor data generates Volume
Typing and User Interfaces
Participatory typing is cumbersome due to small form factor / display keyboard.
Scrolling & Crowded GUIs. Attention issues due to possible mobility. Opportunistic Solutions?
(Location) Privacy
Coarse-grain (cell, wifi) vs. fine (gps)
Energy Consumption
Power Hungry (GPS, Brightness, etc.)

14. Smartphone Crowdsourcing: Challenges
Challenges (Beyond Web Crowdsourcing)
Calibration and Multi-device Issues
Different readings by different sensors (e.g., Wifi RSS, magnetic field, etc.)
Incomplete Data & Quality Issues.
Connectivity Issues
Workforce might have intermittent connectivity (e.g., while travelling) thus can’t provide online readings.
Heterogeneous Clients hinders deployment
Different OSes, sensor, features, APIs, etc.
One supports active background tasks another OS doesn’t, etc.

15. Talk Outline Introduction & Challenges Urban Location Data
Anyplace Indoor Information System
Urban Sensing Testbeds
SmartLab Smartphone Programming Cloud
Urban Trajectory Search
SmartTrace Query Processing Framework
Urban Social Networks
Rayzit Crowd Messaging Service
Objective: To show how some of these challenges have been tackled in our work

16. Urban Location Data
People spend 80-90% of their time inside buildings, while 70% of cellular calls and 80% of data connections originate from indoors.
GPS has low availability indoors due to the blockage or attenuation of the satellite signals but it is also very power hungry.
Smartphones can nowadays localize off-the-shelf with onboard sensors and WiFi signal fingerprints (coined Hybrid Localization)
New Applications:
In-building Navigation (Malls, Airports, Museums, Schools, etc.)
Asset Tracking and Inventory Management (Hospitals, etc)
Elderly support for Ambient and Assisted Living (AAL)
Augmented Reality (Firefighters), Social Networking, etc.

17. Urban Location Data
Indoor Localization using proprietary infrastructure: Infrared, Bluetooth, Visual or Acoustic Analysis, RFID, Ultra-Wide-Band, Wireless Sensor Network, Inertial Measurement Units (IMU), Wireless LAN.
Smartphone Localization:
The Cywee/Airplace Positioning platform performs multi-dimensional (i.e., 3-axis accelerometer, gyroscope and digital compass) motion sensor fusion for calculating the user orientation in real-time and implements an in-house pedometer algorithm for pedestrian trajectory tracking.
Hybrid Localization: Combination of more than 1 techniques such as IMU+WiFi (acceler., gyro, digital compass) MapMatching, Magnetic Data, pedometer

18. I can see these Reference Points, where am I?
Urban Location Data
I can see these Reference Points, where am I?
Cellular
WiFi
(x,y)!
RadioMap Service
...
User u
Cellular

19. Urban Location Data References Cywee / Airplace
[Airplace] "The Airplace Indoor Positioning Platform for Android Smartphones", C. Laoudias et. al., Best Demo Award at IEEE MDM'12. (Open Source!)
[HybridCywee] "Demo: the airplace indoor positioning platform", C.-L. Li, C. Laoudias, G. Larkou, Y.-K. Tsai, D. Zeinalipour-Yazti and C. G. Panayiotou, in ACM Mobisys'13. Video at:
[UcyCywee] IPSN’14 Indoor Localization Competition (Microsoft Research), Berlin, Germany, April 13-14, 2014.
[Anyplace] Crowdsourced Indoor Localization and Navigation with Anyplace, In ACM/IEEE IPSN’14.
Cywee / Airplace

20. Urban Location Data
Anyplace Architecture
Navigator
Viewer, Widget

21. Urban Location Data Live Demo!
Anyplace Indoor Information Service (IIS)
Live Demo!

22. Anyplace Crowdsourcing Challenges
A) Big Data
Massively process RSS log traces to generate a valuable Radiomap
Utilized for KNN positioning
Processing current logs in Anyplace for a single building might take several minutes!
Challenges in MapReduce:
Spatio-temporal Analysis
Missing Values / Outliers / Quality / Multi-device Issues (see next)

23. Anyplace Crowdsourcing Challenges
B) Quality: Unreliable Crowdsourcers, Multi-device Issues, Hardwar Outliers, Temporal Decay, etc.
Remark: There is a Linear Relation between RSS values of devices.
Challenge: Can we exploit this to align reported RSS values?
"Crowdsourced Indoor Localization for Diverse Devices through Radiomap Fusion", C. Laoudias, D. Zeinalipour-Yazti and C. G. Panayiotou, "Proceedings of the 4th Intl. Conference on Indoor Positioning and Indoor Navigation" (IPIN '13), Montbeliard-Belfort France, 2013.

24. Anyplace Crowdsourcing Challenges
C) Privacy
Challenge: How to localize using a Radiomap Service, without revealing my location to the service?
Solution (ongoing): We developed a spatio-temporal privacy scheme using bloom filters coined Temporal Vector Map (TVM).
Provides k-anonymity guarantees
Enables both snapshot and continuous localization.
"Towards planet-scale localization on smartphones with a partial radiomap", A. Konstantinidis, G. Chatzimilioudis, C. Laoudias, S. Nicolaou and D. Zeinalipour-Yazti, ACM HotPlanet '12.
“Privacy-Preserving Indoor Localization on Smartphones with VectorMap”, A. Konstantinidis, P. Mpeis, N. Pelekis, D. Zeinalipour-Yazti and Y. Theodoridis, under submission.

25. Anyplace Crowdsourcing Challenges
TVM Outline:
RadioMap (server-side)
WiFi
...
Bloom Filter (u's APs)
K=3 Positions
User u

26. Anyplace Crowdsourcing Challenges

27. Talk Outline Introduction & Challenges Urban Location Data
Anyplace Indoor Information System
Urban Sensing Testbeds
SmartLab Smartphone Programming Cloud
Urban Trajectory Search
SmartTrace Query Processing Framework
Urban Social Networks
Rayzit Crowd Messaging Service

28. Urban Sensing
Use sensors in urban environments in support of more classic environmental sensing applications.
"People sense and contribute data about their surroundings using mobile devices" (Kanhere)
Example Projects:
Dartmouth | Metrosense: SoundSense, CenceMe, Sensor Sharing, BikeNet, AnonySense, and Second Life Sensor.
MIT | Cartel: VTrack/CTrack, PotHole
Harvard : Citysense (grew out of MoteLab)
UNSW: Noise (Earphone) & Air pollution (HazeWatch, CommonSense),

29. Urban Sensing Monitoring Urban Spaces NoiseMap
"Ear-Phone: An End-to-End Participatory Urban Noise Mapping System " Rajib Rana, Chun Tung Chou, Salil Kanhere, Nirupama Bulusu, and Wen Hu. In ACM/IEEE IPSN 10, SPOTS Track, Stockholm, Sweden, April 2010.

30. Urban Sensing Mapping the Road traffic by collecting WiFi signals.
Received Signal Strength (RSS): power present in WiFi radio signal
Graphics courtesy of: A .Thiagarajan et. al. “Vtrack: Accurate, Energy-Aware Road Traffic Delay Estimation using Mobile Phones, In Sensys’09, pages ACM, (Best Paper) MIT’s CarTel Group

31. Urban Sensing
This kind of a paradigm has nowadays an industrial success.
CrowdSensing app by Waze (Israel) now Google!
Waze: Free GPS Navigation with Turn by Turn
Workers report their GPS location and events (gas prices, traffic jams, etc.)
Real-time updates to users

32. Testbeds
Smartphone Testbeds: Allow the requestor to deploy a task (app, data collection, remote terminal etc.) directly on the end smartphone devices.
[PRISM] T. Das, P. Mohan, V.N. Padmanabhan, R. Ramjee, and A. Sharma, “PRISM: Platform for Remote Sensing using Smartphones”, In ACM MobiSys’10.
[CrowdLab] E. Cuervo, P. Gilbert, B. Wu, and L.P. Cox, “CrowdLab: An Architecture for Volunteer Mobile Testbeds”, In COMSNETS’11.
[PhoneLab] G. Challen et. al. “PhoneLab: A Large-Scale Participatory Smartphone Testbed”, In USENIX NSDI’12 (poster).
[SmartLabDemo] "Demo: a programming cloud of smartphones", A. Konstantinidis, C. Costa, G. Larkou, D. Zeinalipour-Yazti, In ACM Mobisys '12.
[SmartLab] "Managing Smartphone Testbeds with SmartLab", G. Larkou, C. Costa, P. Andreou, A. Konstantinidis, D. Zeinalipour-Yazti, In 27th USENIX LISA '13, Washington D.C. USA, , 2013.

33. Urban Sensing TestBeds
PhoneLab: a Participatory SmartPhone Sensing Testbed (People-Centric Testbed)
200 Nexus S 4G phones used by Students and Faculty Members at the Univ. of Buffalo
Incentive: Free Sprint Phone for 1st year. After that, only $44.23/month for an unlimited plan (claimed to be better than competition)
Targeted for Data Collection Scenarios (not fine-grain access like SmartLab)
Each Data Collection task need to undergo an Institutional Review Board process (similar to other projects touching ethical issues)
Data Collection: Workers (Students) have to bring in their smartphones to have the app installed + data collected.
[PhoneLab] G. Challen et. al. “PhoneLab: A Large-Scale Participatory Smartphone Testbed”, In USENIX NSDI’12 (poster).

34. Urban Sensing Testbeds
We developed a comprehensive architecture for managing smartphones through a web browser.
SmartLab (
40+ Android Devices, Real Sensors, Real Computing Stack
Different Connection Modalities: 3G (unlimited 3G bancwidth by MTN Telecom), Wifi, Wired, Remote 
Static Androids
Mobile Androids

35. SmartLab (http://smartlab.cs.ucy.ac.cy/)
Urban Sensing Testbeds
SmartLab (
Rent
Manage
See/Click
Shell
File Sys.
Automation
Debug
Data
Live Demo!

36. Urban Sensing Testbeds
Scenario I: Data Collection in Smart Cities
How to handle a fleet of Android-powered entertainment equipment installed on 1000 buses?
How to manage a city-scale infrastructure comprising of low-power, low-value Android-oriented devices (installed on traffic lights, etc.)
How to manage a city-scale SETI-like computational cluster comprising of Smartphones.
We tend to change smartphones faster than PCs …

37. Urban Sensing Testbeds
Scenario II: Application Testing (Mockup Studies)
How to test my app automatically on N different smartphones scattered around in a city?
Mockup Sensors
GPS mockup
Accelerometer sensor
Compass sensor
Orientation sensor
Temperature sensor
Light sensor
Proximity sensor
Pressure sensor
Gravity sensor

38. Urban Sensing Testbeds
Scenario III: Personal Gadget Management
How to manage my personal gadgets at a fine-grain (i.e., clicks, file-transfer, update, etc.)
eReaders
Rasperry PI
Tablets
Smart Watches
Smart TVs
SmartBooks
Smart Glasses
Smart Home Phones

39. Talk Outline Introduction & Challenges Urban Location Data
Anyplace Indoor Information System
Urban Sensing Testbeds
SmartLab Smartphone Programming Cloud
Urban Trajectory Search
SmartTrace Query Processing Framework
Urban Social Networks
Rayzit Crowd Messaging Service

40. Urban Trajectory Search
Fact:
Smartphones can collect positional (x,y) in a power efficient manner (e.g., iPhone triangulated log file, Android Geolocation wardriving).
Crowdsourcing Incentive:
Contribute to the resolution of queries for Social Benefit (without revealing traces).
Applications:
Intelligent Transportation Systems: “Find whether a new bus route is similar to the trajectories of K other users.”
Social Networks: “Find if there is an evening cycling route from MOMA to the Julliard”
GeoLife, GPS-Waypoints, Sharemyroutes, etc. offer centralized counterparts.

41. Urban Trajectory Search
Problem: Compare a query with all distributed trajectories and return the k most similar trajectories to the query.
Similarity between two objects A, B is associated with a distance function.
D = 7.3
D = 10.2
D = 11.8
D = 17
D = 22
Distance K ?
Query

42. Urban Trajectory Search
An intelligent top-K processing algorithm for identifying the K most similar trajectories to Q in a distributed environment.
Step A: Conduct an inexpensive linear-time LCSS(MBEQ,Ai) computation on the smartphones to approximate the answer.
Step B: Exploit the approximation to identify the correct answer by iteratively asking specific nodes to conduct LCSS(Q, Ai).
"Crowdsourced Trace Similarity with Smartphones", Demetrios Zeinalipour-Yazti and Christos Laoudias and Constandinos Costa and Michail Vlachos and Maria I. Andreou and Dimitrios Gunopulos, IEEE TKDE, Vol. 25, , 2013.
"SmartTrace: Finding similar trajectories in smartphone networks without disclosing the traces", Costa et al., IEEE ICDE'11.

43. Urban Trajectory Search
SmartTrace for Android (open source)!
Query Q
Device B
Device C

44. Urban Trajectory Search
Privacy Setting
Answer
With Trace
Answer

45. Talk Outline Introduction & Challenges Urban Location Data
Anyplace Indoor Information System
Urban Sensing Testbeds
SmartLab Smartphone Programming Cloud
Urban Trajectory Search
SmartTrace Query Processing Framework
Urban Social Networks
Rayzit Crowd Messaging Service

46. Urban Social (Crowd) Networks
Social Media (Facebook, Linked-in, … ) utilize a Social Graph (friendship, follower, followee) to map the relationships between users.
Social Media in Urban Settings: Issues
Urban Applications many times require location-based rather than social-based interactions, e.g.,
Inform my neighboring drivers about an accident (e.g., in Waze).
Inform people in a city about an event.
Location-based services suffer from bootstrapping
e.g., Check in to Foursquare and find nobody else there
Interacting with the Crowd, calls for stronger Privacy!

47. Urban Social (Crowd) Networks
We developed Rayzit for Windows Phone after receiving an Industrial Award by the Appcampus Program (Microsoft, Nokia & Aalto, Finland).
Ranked among the 5 best apps of the given program among 3500 submissions.
A few thousand downloads and active users on our big-data backend.

48. Urban Social (Crowd) Networks
Find 2 Closest Neighbors for ALL User
"Continuous all k-nearest neighbor querying in smartphone networks", Georgios Chatzimilioudis, Demetrios Zeinalipour-Yazti, Wang-Chien Lee, Marios D. Dikaiakos, In IEEE MDM'12.

49. Crowdsourcing Urban Data with Smartphones
Demetrios Zeinalipour-Yazti
Data Management Systems Laboratory
Department of Computer Science
University of Cyprus
Thanks!
Questions?
Invited Talk at the Mining Urban Data (MUD) Workshop
(with EDBT/ICDT), Athens, Greece, March 28, 2014