Four of the five members of Team Wibble. From left to right: Max Glenister, Carl Marshall, Martin Rosenmejer & Sjors Timmer. The picture was taken by the fifth member, Kevin Carmody.

Team Members

Max Glenister | @omgmog, Personal website, Blog

Martin Rosenmejer | @rosenmejer

Kevin Carmody | @skinofstars

Carl Marshall | @cml3net

Sjors Timmer | @sjors

Initial brainstorm

Sketching out what we do on our communte

Audience

• us
  • our likes
  • current usage
    • apps
      • gmail
      • twitter
      • facebook
      • bbc news
      • email
      • sms
      • web browsing
      • british rail
      • IM (gtalk, etc)
    • book/music
      • kindle
      • physical books
      • ipod/etc.
    • mobile/smartphone/tablet
      • iOS
      • Android
      • Blackberry

Getting the ideas up there

Logistics

• safety
  • "help I've run over someone!"
  • unforeseen delays/accidents
• connectivity
  • data coverage
  • "always on" internet
  • work around bad coverage
    • improved coverage
    • caching
    • mesh networking
• standing/sitting
  • is the bus busy
  • can we get a seat
• changing mode of transport (going from bus to walking for example)
  • route optimisation
• location tracking
• functions
  • time of day affects function
    • AM - productive/focussed
    • PM - relaxing/consuming

Everyone gets a dot to vote on ideas they like.

Entertainment

• music
• podcasts
• reading
  • escape
  • study
• news
• social
  • twitter
  • facebook
• email

Sorting the ideas into common groupings

Attributes

• pre-emptive
  • get data in advance
• passive or active
  • push notifications
  • refresh/sync
• voice driven
  • audio commands
  • audio feedback
  • flexibility - able to switch between text/audio to suit situation
• bite-size/consumable chunks
  • segmented
  • easy to consume
  • not sensitive to distraction

The idea

Challenges/features

• Geolocation - privacy - anonymously submitted information, not socially oriented
• Quality of input - controlled criteria for rating a journey - multiple choice, e.g. "Noise - 1-5", ability to leave an additional comment
• Account persona - when setting up account for app you determine the kind of commuter that you are, e.g. student, business professional, pensioner, etc.
• Frequency of people in area - see "hot spots" on google maps, filterable by type of person (See personas)

Needs/context

• context
  • simple
  • consuming
  • last minute changes for speed and/or comfort
  • commuter generated information

• commute optimisation within public transport by speed and comfort
• is it clean?
• can I drink my coffee on the next bus
• I want to read a book on the bus
• will it be quiet?
• will it be filled with noisy school children
• what should I wear?

The Design

In a nutshell

This is a mobile device application to aide the commuter when picking a route for a common "run" allowing them to select on factors that address comfort, as well as time and cost.

Assumptions

• People will do what they want to do on their commute, our job is to take away the surprises of the journey and let them find a route that suits their wants.
  • We're not trying to entertain them
  • We want to be a quick reference
• This is a mobile device application, so it is device/platform specific in appearance, but also because it has to work whilst "offline"
• Monetizing is beyond the scope of this activity, but we recognise that the data collected could be of high value to both transport services and some marketing.
• Some data needs to be created by the userbase, so there will be grow issues for some features early on.
• We are focussed on the following modes of transport:
  • Bus
  • Train
  • Tram
  • Tube
  • NOT: Car, Walking or Bike

Use Cases

1) Initial app registration/persona building 2) Tutorial / guide 3) First check of the journey 4) Review something / instant survey 5) Setup commute points + times 6) Edit profiles/persona

Factors for routing on

x = external data

CORE Time

• Speed of journey x
  • Traffic flow x
  • Number of connections x
  • Cost! x

Comfort

• Noise Levels -- ????device microphone sampling?
• Space
  • Sitting - counting spare seats?
    • Legroom - datasets based on vehicle type
    • Headroom - ditto
• Connectivity
  • Phone signal - dataset and automated sampling
  • Wifi - dataset and automated sampling
  • Power socket - vehicle type
• Facilities
  • Toilet - dataset + user rating

ADDITIONAL What is comfort?

• Smell - u
• Space
  • Luggage - combine vehicle and user
  • Cycle storage - combine vehicle and user
  • Pushchair - combine vehicle and user
  • How much bumping around - dataset
  • driving style - rate the driver?
• Lighting
  • Windows? - time of day, vehicle type, windows clean?
  • Artificial
• Amount of connections - dataset
• Facilities
  • Food - dataset + user rating
  • Drink - dataset + user rating
    • Alcohol - dataset + user rating + event information?
      • for consumption?
      • being consumed?
• Temperature

Data Sets

These data sets can be used to help the routing engine determine timings. These will need to be supplemented by various user generated data sets.

Trends Data

• NPTDR - national transport snapshot survey - http://data.gov.uk/dataset/nptdr
• average transport flow rates flow - http://www.transport-watch.co.uk/transport-fact-sheet-1.htm
• travel surveys -
• transport trends - http://data.gov.uk/dataset/transport_trends
• Bus reliability trends - http://data.gov.uk/dataset/quarterly_bus_statistics_great_britain
• Road Conditions - http://data.gov.uk/dataset/road_conditions_in_england

Real Time Traffic

• GB Traffic Matrix - Annual average daily flow estimates for locations on the Motorway and A road network (interactive site). http://data.gov.uk/dataset/gb-traffic-matrix
• NAVTEQ Real time traffic

Transport Timetabling

• Public Transport Access Nodes - http://data.gov.uk/dataset/naptan

Events/Probable People Density

• Eventful API - http://api.evdb.com/

Routing Engines

• Google API provides minimal routing restriction. You can add 8 waypoints or use there are two possible exclusions (tolls & highways)

Others Of Interest

• Tube Exits app - http://www.tubeexits.co.uk/

User Generated Data

• Noise Levels - we considered having the application use ambient noise level sampling to help this, but think it is unlikely to be practical
• Empty Seat Counting - surveying users to count empty seats
• Phone and Wifi signal availability

MISC NOTES

Presentation (video)

Watch the interim and final presentation of Team 3 Team Wibble:

Video Team 3