Quality of Experience in Smart Environments
0 likes337 views
The document presents a framework for enhancing the quality of experience in smart environments, particularly in public transportation. It discusses the integration of user and environment data through a cloud-based platform called Cloud2Bubble, which aims to improve user interaction and satisfaction. The ongoing field experiment in Porto, Portugal, collects user feedback and environmental data to assess and enhance the public transport experience.
Quality of Experience in Smart Environments
- 1. Quality of Experience in Smart Environments Pedro Maurício Costa DEGI Club, FEUP Intelligent Systems and Networks Porto, 8 June
- 2. Introduction Background Context • Smart Environments • Human Computer Interaction • Quality of Experience Proposed Framework • Cloud-based software platform • Device integration Domain of Application: Public Transport Field Experiment Conclusions
- 3. Background 2007 • Licenciatura em Engenharia Informática e Computação 2007 - 2009 • Software Engineer at Critical Software 2009 - ? • PhD, Mobile Human Computer Interaction • Jeremy Pitt • João Falcão e Cunha • Teresa Galvão
- 4. Context Pervasive Computing • miniaturisation of devices • modern communication networks • Internet of Things Human Computer Interaction • shift in paradigm from PCs to multiple devices • Affective Computing Smart Environments • making human interaction with the system a pleasant experience Quality of Experience • “the degree to which a system meets users’ expectations”
- 5. Data Flows Data streams • user and environment generated • app and service offering: excessive choice may be detrimental and frustrating Collection of Data Service Offering 1M+ apps available (iOS + Android)
- 6. Siri Intelligent personal assistant • natural language interface • personalises results • adapts to user’s individual preferences • curator of things that are relevant Concerns developing Siri include: • gender neutral? • being friendly and humble • occasional light attitude • “How would we want a person to respond?”
- 7. Affective Computing A novel dimension of interaction • to make systems more emphatic • implicit interactions with the system » camera » microphone » patterns of usage Activation » physiology measurement Nervous Excited Model for emotion representation • Russell’s circumplex • valence (x): cognitive Unpleaseant Pleaseant • arousal (y): physiological Depressed Relaxed Deactivation
- 8. Framework Addresses the disconnection between two main data streams Continuous loop of interaction between user and system • leverages existing technology and context • integrates emotion Collection of user and environment data • world and user representation Delivery of personalised services • enhancing Quality of Experience • influence user behaviour User-centric and systemic
- 9. Cloud2Bubble Cloud 2 Bubble abstract computing infrastructure user-centric, local *-as-a-Service composed of devices and services aggregates data streams collects user and environment data generates individual profiles delivers relevant services action generation and selection
- 10. Cloud2Bubble Smart System • collects user and environment data • assesses its current state • generates and selects actions • acts on the environment World Modelling • high-level events Decision Making • environment model • user profiling World Modelling Decision Making • imprecise and incomplete data Hardware • fuzzy reasoning Environment Action Perception Execution
- 11. Public Transport Domain of Application • Public Transport: Intelligent Traveller Information Systems • poor Quality of Experience • equipped modern networks (sensors, info) • large user base with idle time • benefits for both users and providers
- 12. Mobile Device App for Public Transport Establish connection between users and the system • collects user and environment data • pre-processes data • requests feedback • receives notifications • integrates with the environment MSc project: João Vieira
- 13. PTSense prototype Non-functional prototype, high-fidelity • based on the main features identified » collect data & deliver service • Android platform » large user base » ease of development • mobile sensing platform • usability testing in London Functional prototype • available on Google Play • sub-set of features: purely participatory; public transport independent • in use for the on-going experiment
- 14. Field Experiment Deploy the system on a real-world environment: Porto, Portugal stage 1: assess Quality of Experience • Experience Sampling Method-based • collect user and environment data • correlation between environment and user feedback • little intervention required from PT providers stage 2: enhance Quality of Experience • delivery of services, identified previously • requires tighter integration with PT providers
- 15. Collection of Environment and User Data •personal •environment •feedback
- 16. Assessing Quality of Experience Acquiring users, Porto Portugal • survey for selecting users: 170 replies • demographics • travelling habits • smartphone OS and usage (limited to Android users) • “I'm willing to actively contribute with daily, anonymous reports on my Public Transport experience through my smartphone during the course of this experiment”
- 17. “How can your journey be improved?” Comfort 16% Real-time info Trip planning 18% 8% Price 5% Internet 9% Entertainment 5% Service Nothing 32% 8%
- 18. Other replies “Eu não me importo com os atrasos, desde que seja avisado” “Quero saber se posso ficar mais 5 minutos a tomar o meu café ou se tenho de me apressar” “Temperatura mais agradável no metro, é preciso levar casaco” “Viagens grátis com bebidas grátis a caminho do piolho (:D)” “O ar é abafado no autocarro quando está demasiada gente”
- 19. On-going in Porto, Portugal User Sample • 10 regular travellers • 18 - 35 years old • students and professionals • higher education Goal • collect environment data and user feedback: » 2 weeks, 10 working days » 2 journeys / day » total of 200 journeys Running until mid June • incidents related with heterogenous version and devices • some of the participants quit after confirmation • so far: 110 journeys, 25+ hours
- 20. Contribution Forward thinking, user-centric and systemic Generic smart framework Body of knowledge: Quality of Experience Usage of emotion Instantiation in public transport Opens up a number of opportunities
- 21. Workshop on Quality of Experience in Mobility Contexts 10 September @ hci2012, Birmingham UK submission deadline: 15 June www.mqoe.org
- 22. References • Cloud2Bubble: a context-aware platform for enhancing Quality of Experience, in awareness magazine, 2012; • Cloud2Bubble: Enhancing Quality of Experience in Mobile Cloud Computing Settings, in The Third ACM Workshop on Mobile Cloud Computing and Services, 2012 • Smart Mobile Sensing for Measuring Quality of Experience in Urban Public Transports, in The Second International Workshop on Smart Mobile Applications, 2012 Cloud2Bubble www.cloud2bubble.com Pedro Maurício Costa pm.costa@imperial.ac.uk www.pedromauricio.com