more accurate bus prediction allows passengers to find alternate forms of transportation do this...

13
Predicting Bus Arrival Time with Mobile Phone based Participatory Sensing

Upload: natalie-barron

Post on 29-Mar-2015

212 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: More Accurate Bus Prediction Allows Passengers to find alternate forms of transportation Do this with energy efficiency in mind Dont use any high level

Predicting Bus Arrival Time with Mobile Phone based Participatory

Sensing

Page 2: More Accurate Bus Prediction Allows Passengers to find alternate forms of transportation Do this with energy efficiency in mind Dont use any high level

Goal

More Accurate Bus Prediction› Allows Passengers to find alternate forms

of transportation Do this with energy efficiency in mind Don’t use any high level permissions

Page 3: More Accurate Bus Prediction Allows Passengers to find alternate forms of transportation Do this with energy efficiency in mind Dont use any high level

Equipment Used

Microphone – Record Sound Cell Signal – Determine Location Accelerometer - Determine Bus or

Train

Page 4: More Accurate Bus Prediction Allows Passengers to find alternate forms of transportation Do this with energy efficiency in mind Dont use any high level

System Design

Page 5: More Accurate Bus Prediction Allows Passengers to find alternate forms of transportation Do this with energy efficiency in mind Dont use any high level

System Design Cont.

Query User – Looks for Bus arrival time by indicating bus route and stop

Sharing User – Contributes mobile sensing information to the backend server› Information includes – a collected cell sequence

from nearby cell towers, sound and accelerometer data to make sure the user is on a bus

Backend Server – Processes data from sharing users and give information to querying users

Page 6: More Accurate Bus Prediction Allows Passengers to find alternate forms of transportation Do this with energy efficiency in mind Dont use any high level

Backend Server Data

Maintains a database of sequences for cell tower IDs for the different Bus routes

Page 7: More Accurate Bus Prediction Allows Passengers to find alternate forms of transportation Do this with energy efficiency in mind Dont use any high level

On a Bus?

Sound detection

Page 8: More Accurate Bus Prediction Allows Passengers to find alternate forms of transportation Do this with energy efficiency in mind Dont use any high level

Bus vs Train

Accelerometer Readings

Page 9: More Accurate Bus Prediction Allows Passengers to find alternate forms of transportation Do this with energy efficiency in mind Dont use any high level

Bus Classification

Sequence Matching After running an

Algorithm the Serverdetermines which route has the bestscore and that determines what bus the sharing user is on

Page 10: More Accurate Bus Prediction Allows Passengers to find alternate forms of transportation Do this with energy efficiency in mind Dont use any high level

Arrival Time Prediction

After all data is uploadedand each bus is determined where it isAny querying user will be able to get data onwhere the bus is andapproximate arrival time.

Page 11: More Accurate Bus Prediction Allows Passengers to find alternate forms of transportation Do this with energy efficiency in mind Dont use any high level

Findings from Experimentation

Page 12: More Accurate Bus Prediction Allows Passengers to find alternate forms of transportation Do this with energy efficiency in mind Dont use any high level

Findings Cont.

Page 13: More Accurate Bus Prediction Allows Passengers to find alternate forms of transportation Do this with energy efficiency in mind Dont use any high level

Limitations

No Users on a Bus› Causes bus times to be reported wrong

Overlapped Routes › The Server will sometimes misinterpret a

route