A Comfort Measuring System for Public Transportation Systems Using Participatory Phone Sensing

1
A Comfort Measuring System for Public
Transportation Systems Using Participatory Phone
Sensing
Cheng-Yu Lin1, Ling-Jyh Chen1, Ying-Yu Chen1, and
Wang-Chien Lee2 1Academia Sinica, Taiwan 2The
Pennsylvania State University at University Park,
USA
2
What are people doing on the bus?
Comfort does matter!!
3
How to measure it?
Questionnaire/Interview
Professional Instruments
Problems Cost, Timeliness, and Scalability
4
Participatory Phone Sensing

• A new sensing paradigm to exploit the sensing
  capabilities of modern smart phones to gather,
  analyze, and share local knowledge of our
  surroundings (e.g., CenseMe, SoundSense,
  Nericell)
• It does not rely on dedicated sensing
  infrastructures and the top-down model of data
  collection.
• It is more penetrative, and encourages
  participation at personal, social, and urban
  levels.

Question how about lets combine the
participatory phone sensing and top-down data
collection model?
5
Comfort Measurement System

• Goal to evaluate the comfort level of public
  transportation systems

Public Transportation Systems
Participants
Sensing data (e.g. locations, acceleration, and
time)
Authorized data (e.g. bus trajectories and
vehicle properties)
Data Mashup and Statistics
Scoring and ranking results
6
Our Contributions

• We propose the Comfort Measurement System that
  exploits participatory phone sensing (bottom-up
  model) and the authorized data (top-down model).
• We prototype a CMS, called TPE-CMS, to evaluate
  the public bus transportation service in Taipei
  City.
• We conduct a 70-day experience to reveal the
  insights of the Taipei e-bus system.

7
Phone Sensing

• Exploit the GPS and G-sensor (3-axis
  accelerometer) of modern smart phones
• Calculate comfort index by following ISO 2631

8
Authorized Data

• No need to reinvent the wheel!
• We take advantage of existing real-time bus
  tracking systems, which are available in many
  major cities world-wide (e.g., Boston, Cambridge,
  Seattle, and Taipei).
• It contains the bus trajectory, route number,
  operating agency, and the other useful data.
• This may be the most challenge, because you have
  to talk to the authority ?

9
Data Mashup
Bus Trajectory
User Trajectory
We suppose the user is on the b-th bus, s.t. b
arg Min Di
10
Implementation
4,028 buses, 287 routes, 15 agencies, and 1
sample per minute
http//VProbe.org/TPE-CMS/
11
Experiments

• Period 2010/03/15 2010/07/22
• 15 volunteers
• Collect trajectory and vibration traces of Taipei
  buses using Android phones
• Keep a memo of the ground truth (i.e., the
  agency, route, and license number of their bus
  rides)
• 425 trajectories collected, involving 12 agencies
  and 3 types buses

12
Results(1/3) - Trajectory Matching Results
13
Results(2/3) -The Statistics based on Buses Types

• Light buses are uncomfortable.
• No significant difference between the standard
  buses and the low-floor ones.

14
Results(3/3) - The Statistics based on Buses
Agencies

• The most comfortable and uncomfortable agencies
  are exactly the same as the ones reported in the
  survey made by Taipei Department of
  Transportation.

15
Conclusions

• We present a Comfort Measuring System for public
  transportation systems, and prototype the system
  in Taipei city.
• The CMS system can be deployed in any cities, as
  long as there are volunteering participants and
  there are authorized transportation data
  available.
• Work on analyzing other factors that affect
  comfort levels is ongoing (e.g., road conditions,
  drivers behavior, and traffic congestion).

16

• Thanks for Your Attention!

http//VProbe.org/ http//VProbe.org/TPE-CMS/