The Smart Party: A Personalized Location-aware Multimedia Experience

1
The Smart Party A Personalized Location-aware
Multimedia Experience

• Kevin Eustice, V. Ramakrishna, Nam Nguyen, and
  Peter Reiher
• CCNC
• January 11, 2008

2
The Ubicomp Home of the Future

• Increasing numbers of computing and
  communications devices in the home
• Wireless networks are common in the home
• The family and visitors carry personal devices
  with them
• What new things can we do for people with these
  capabilities?

3
One Example The Smart Party

• People commonly carry devices that hold the music
  they like
• Speakers can be connected to the home network
• Localization techniques can determine where
  people are in a house
• So why not combine those abilities to improve a
  party in the house?

• Provide a dynamic customized music mix to
  match the tastes of people in different rooms in
  the house

4
What Weve Got

5
What We Want
6
The Basic Approach

• Run a ubiquitous computing application
• On both devices in the home and personal devices
• Built using Panoply middleware
• Automatically recognize and enroll people you
  invited to the party as they arrive
• Figure out everyones location
• Use musical preferences of those in a room to
  choose music
• Get chosen music from personal device and play it

7
Major Challenges

• Configuring the personal devices
• Only for those invited, of course
• Localization
• Selecting the music
• To maximize overall satisfaction of partygoers
• Handling dynamism

8
Smart Party Architecture
A partygoer arrives
The personal device localizes
Access Negotiation
Smart Party
Localization Map Updates
Mary SmartParty UserDeviceApp
Bob
Family Room SmartPartyLocationApp
Living Room SmartPartyLocationApp
A room gathers music preferences
Music is selected, fetched, and played
Kitchen SmartPartyLocationApp
Content Provisioning Protocol
Content Selection Protocol
Sam SmartParty UserDeviceApp
Alice SmartParty UserDeviceApp
9
Handling the Hard Parts

• Configuration
• Cryptographic voucher carries both invitation and
  necessary configuration information
• Localization
• Using maps of 802.11 wireless characteristics
• Selecting and retrieving the music
• Automated voting algorithm based on known user
  preferences
• Handling dynamism
• Middleware builds and maintains groups based on
  users locations

10
Smart Party Status

• Working version in UCLA LASR lab
• Based on our Panoply ubicomp middleware
• Required around 1800 lines of Java
• Continuing work
• To measure user satisfaction
• To study different methods of choosing music
• To handle changing wireless maps
• New application features

11
Conclusion

• The Smart Party is an example of neat things
  ubicomp can do for users
• The Panoply middleware made the Smart Party easy
  to build
• The Smart Party demonstrates the value of
  organizing applications by groups
• The Smart Party demonstrates practical solutions
  to common ubicomp problems