Part Two: Initial Prototype and Evaluation Plan
Version 2.0 (11/23/97)

Project Team:

Jason Elliott
Kevin Scott
Alexandre Stoytchev
Rawesak Tanawongsuwan

Send Us E-Mail




Project Description

We plan to implement a home audio system which is capable of directing sound to specific speakers throughout the home based on the user's location, creating the impression that the music is following the user. In addition, the system will be able to adjust the stereo settings based on the user's personal preferences and current situation or location. The project will incorporate indoor positioning technology with ubiquitous interface design and existing audio technology to produce a useful and novel addition to the home of the future.

Our concept of "intelligent home applications" extends beyond audio. We envision work areas, television programs, and communication media that can follow the user throughout the home as well. To narrow the scope of our project, we chose audio as an example of such an "intelligent home application." Since these applications will be used in the home, they need to be designed with a broad user population in mind. We must attempt to accommodate as many age levels and disability levels as possible. One way we have approached this problem is by allowing a voice interface to the system in addition to a touch screen with a GUI. The voice interface extends the usability of the system to include users who are illiterate (especially young children) as well as users who are not able to reach out and manipulate a graphical interface.

  • What exactly will the Follow Me system do?

  • How will the Follow Me system accomplish these things?

    Design Criteria

    FollowMe will be evaluated in a lab setting, with the assistance of two members of the design and development team. The environment will be a simulated multi-room home, with a desktop computer providing a prototypical wall mounted or handheld interface. In addition, a voice interface will be simulated. The system will be evaluated by the development and evaluation teams on the basis of the following criteria:

    Robustness

    Accessibility

    Satisfaction

    Description of New Design


    Link to sketches and other resources.

    The primary visual interface for FollowMe has undergone three major revisions. The first sketches were produced during a brainstorm session, attempting to stretch our concept beyond today's limitations, to a truly ubiquitous home audio control system.

    The second first major revision was produced for an in-class exercise in Cognitive Walkthroughs. These simple graphics were generated to show how a user might add a song to their audio playlist. We (the design team) learned quite a bit from that in-class experience, and subsequently incorporated a number of suggestions into the final design revision.

    The final revision will be used during the team evaluation process. It consists of a number of web hosted storyboards that are linked together to demonstrate basic interface functionality.

    Description of Design Prototype

    Prototype of Touch Screen Interface
    Layout of the Prototype in the CPL
    Camera Perspectives
    Positioning System

    The FollowMe prototype models a home audio system which is capable of directing sound to specific speakers throughout the home based on the user's location, creating the impression that music is following the user. In addition, the prototype demonstrates the concept of adjusting the stereo settings based on the user's personal preferences and current situation and location. The prototype incorporates an indoor positioning system to locate a user in a virtual room. The system utilizes two cameras that capture images in the virtual home and calculates the location of the user in room coordinates. All image processing is performed on two Silicon Graphics Indy workstations. FollowMe also utilizes the built-in audio capabilities of these two workstations, and along with a third workstation, maps each machine's audio output capabilities to a room in the virtual home. The audio control and positioning software communicate via a streams based protocol. Based on the user's position in the virtual home, each workstation's speaker is faded up or down, establishing the illusion that music follows a user through the home.

    While this software/hardware combination provides most of the basic functionality we require, there is also a need for a robust and flexible voice recognition system to allow more complete interaction. To this end, an design team member will be available to "implement" this voice recognition interface, interpreting user requests of the system. This will allow the design team to more closely mimic how the system would function in a home of the future.

    Finally, our indoor positioning system must be calibrated before use. During this calibration process, everyone will be asked to step outside the lab space so that the camera can capture the background information it needs to successfully track a subject. Once this calibration step is complete, evaluators will be invited back into the lab, one-at-a-time, to experience the system. While this is obviously not an ideal working environment, at present our prototype is limited to tracking only one subject.

    Evaluation plan

    The three evaluation exercises that we are going to use are: think aloud, cognitive walkthrough, and questionnaire. For each exercise we set up different objectives and require the evaluation team to do different tasks. Since we plan to support alternative user input (voice and touch screen)we split these two forms of communication between the first two evaluations. This will simplify considerably the task of the evaluation team since they will not be confused about which form of input to use. However, the tasks that they will be asked to do will not be totally different in their nature so that comparable feedback about the usefulness and efficiency of the two interfaces can be obtained.

    Think aloud. In this part the evaluation team will use voice interface only. They will be required to perform several routine tasks and requested to explain each step they do in detail. By doing this evaluation exercise we hope to gain some insights into the different kind of tasks the users might be willing to carry out using the system.

    Cognitive Walkthrough. This method is very useful in the early stages of the design process. Since we did not have enough time to develop a fully functional prototype it can be extremely helpful to identify problems with the design of the system and its interface. We did a similar exercise in class and it helped us find some conceptual errors in our design. For example, the users wanted to play a song immediately while the only option that we were providing was to add a song to the play list (queue) which might or might not be empty. In this part the evaluation team will use touch screen interface only.

    Questionnaire. We decided to use a questionnaire to evaluate the concepts and the ideas behind the system rather than the interface. Since this is an unique system we would like to know whether the users like it or not. The implementation details and the prototype are not that important here and we kindly ask the people that are doing the questionnaire to ignore them. Instead we require them to think in abstract terms about the usefulness and viability of the system. We also ask questions about the social impact of the system and whether it is attractive for home use.

    I. Think Aloud

    The think aloud will be performed in the CPL lab on the 3rd floor of CCB

    Think Aloud is a way to gather information by observing users' interaction with a system. While performing a Think Aloud evaluation of the FollowMe system, evaluation team members will be asked to complete a set of pre-determined tasks. As the team members work with the system, they will be asked to provide stream-of-consciousness feedback. In other words, the evaluators will be asked to talk out loud about what they are thinking while using the FollowMe system. The evaluators should elaborate by describing what they believe is happening with the system, why they take action, and what they are trying to do. We will request that their comments not be inhibited in any way, but rather that they be frank and honest.

    There will be two design team members present to assist with the evaluation process, and the process will be recorded via videotape for review by the design team. The system designers will use this evaluation experience to help them further refine interface components and to gauge the relative comfort level of the evaluation team during their exposure to a heretofore unseen technology.

    FollowMe, in its prototypical form, demonstrates most of the functionality that one would expect from a complete system. As mentioned previously, the Think Aloud exercise will stress vocal and kinesthetic interaction as opposed to a purely visual mode.

    In doing so, the design team hopes to create a realistic interactive environment: one that allows evaluators to concentrate on their interaction with the system, rather than fumbling with it's prototypical implementation.

    What follows is a task story that will provide evaluators with both motivation and instructions for their interaction with FollowMe:

    You live in a wonderful new house equipped with a startling new technology called FollowMe. Your home has many rooms, each capable of sensing your motion, and each capable of reproducing your favorite musical selections. Each room is also equipped with touch sensitive displays and voice recognition technology so that you can communicate to your home in a variety of ways.

    You begin your experience in the living room, after a long day at work. While relaxing in your favorite chair, you ask FollowMe to play your favorite song by saying...

    1) "turn on the audio system"
    2) "play the first song from Enya's most recent album"
    3) "volume up two levels"

    While you close your eyes to enjoy your favorite song, your stomach begins to growl - growing louder until it begins to drown out Enya's haunting rhythms. You get up from your chair to grab a bite to eat from the kitchen. As you leave the living room, it's speakers fade out and the kitchen grows louder to your preset volume. The only problem is, the dishwasher happens to be running and you can't hear a thing. You adjust the volume again by saying...

    4) "volume up two levels"

    After fixing a sandwich, and downing a cold glass of milk, you retreat to your bedroom to change clothes. As the kitchen speakers fade away, the bedroom system activates. With a full tummy and more comfortable clothes, your mood changes - looking for something a bit more up-tempo. You know there is a fast song that Enya plays - but you just can't remember the name of it. You ask the system to give you a preview of the songs from Enya's album "In Memory of Trees"

    5) "preview all the songs in Enya's Memory of Trees album"

    After you listen to the first few previews, the song you were thinking of begins to play.

    6) "play this song now"
    7) "set the volume down one level"

    You venture back into the living room and lay down on the sofa. Your mind drifts off... Now it's after midnight and you're ready for bed. You head back to the bedroom. So you can sleep in peace and quite, you tell FollowMe to leave you alone for a while.

    8) "Don't follow me here - it's after midnight"

    In the future, FollowMe will remember that it shouldn't play music in your bedroom after midnight. As you drift off to sleep, FollowMe automatically turns itself off to conserve energy, awaiting a new day.

    II. Cognitive Walkthrough

    Intended Users General Population with the following assumptions/conditions:
    • Users are expected to have some domain knowledge (they should know what music is and know common home audio terms such as "CD", "volume", "track", "play", "stop")
    • In order to use the touch screen with GUI, users should understand what a touch screen is as well as the basic concepts of a GUI (buttons, selection lists, etc)
    • Illiterate population (including small children) are not expected to use the touch screen with GUI
    • Mute population (including very small children) are not expected to use the voice interface
    • Hearing impaired users will not benefit from this system (however, they may benefit from other applications which use the "follow me" concept)
    • Physically disabled users may not be able to control the touch screen with GUI
    Description of the Project See Project Description above.
    Task Descriptions Task I: Set Preferences
    The user should be able to set the preferences for volume level and following status in each room of the house.

    Task II: Play a Song Now

    The user will choose a song that will be played as soon as it is chosen, as opposed to adding the song to the play list or queue. The chosen song might interrupt whatever is currently being played by the system.
    Action Sequences for Tasks Task I: Set Preferences
    (For this evaluation, print this form and then start by clicking here)
    1. (Click "Follow Me" on the Splash screen to begin)
    2. Click on the Audio button
    3. Click on the Preferences button
    4. Click on the Living Room in the floorplan
    5. Turn on Follow Me by clicking on the Follow On radio button
    6. Turn up the volume by clicking on the right arrow beside the volume meter
    7. Click "Done"
    8. (Click "Exit" to close the window)


    Task II: Play a Song Now
    (For this evaluation, print this form and then start by clicking here)

    1. (Click "Follow Me" on the Splash screen to begin)
    2. Click on the Audio button
    3. Click on the button to Pick a Song Now
    4. Click on the Bookmarks button
    5. Click on the song "Too Much" by Dave Matthews Band
    6. Click "Done"
    7. (Click "Exit" to close the window)

    III. Questionnaire

    The designers of this system came up with this questionnaire in order to asses the "follow me" idea. Since we are not aware of any existing system similar to this one we want to get some feedback from you about the usefulness of our system. Please do not pay too much attention to the existing system and interface. We want you to think in abstract terms and help us evaluate the viability of the concepts behind the system.

    Printable Questionnaire

    Please read all questions carefully and try to answer all of them. We thank you in advance.

    1. Your age:

      a) 12-16
      b) 16-18
      c) 18-21
      d) 21-25
      e) 25-35
      f) 35-45
      g) 45-60
      h) above 60

    2. Sex:

      a) male
      b) female

    3. Education:

      a) primary
      b) high school
      d) Bachelor's Degree
      e) Master's Degree
      f) Doctoral Degree

    4. What was your first impression about the system:

      a) I liked it a lot.
      b) I though that it is a crazy idea but I started to like it soon.
      c) I was not surprised at all.
      d) I was disappointed by the limited functionality of the system.
      e) I did not like it at all.

    5. Would you like to use such a system?

      a) Definitely.
      b) Probably.
      c) I'm not sure.
      d) Probably not.
      e) No.

    6. Would you install it in your home?

      a) Yes, when do you release the first version.
      b) Probably yes, but it depends on the price.
      c) I'm not sure.
      d) Probably not, unless it is free.
      e) No way. I still have some brains left.

    7. Did the system make you change your listening habits?

      a) Yes, I almost forgot how to operate a normal CD player.
      b) Yes, I found myself doing things that were not possible without this system.
      c) A little bit.
      d) I'm not sure.
      e) No.

    8. Do you like the idea of the sound following you around the house?

      a) Yes, I always wanted my stereo to be able to do something like that.
      b) It is kind of cool but I'm not used to it.
      c) I'm not sure.
      d) It creates more problems than benefits.
      e) No.

    9. Can you think of any alternative uses of the concept of "follow me"?
    Can it be applied to something that is not sound? Give examples if possible.









    10. Do you think that the system may cause any social conflicts if you were to install it in your house?

      a) Yes, my family will go crazy.
      b) Yes, I will disturb everybody in the house.
      c) I can't tell, I have to install it first.
      d) No, if the settings are done correctly.
      e) No, it will actually make my family happier.

    11. Is it going to make you feel better in your home?

      a) Yes, I'd love to tell the system just before I go to bed "Play my favorite classic song".
      b) Yes, now I can listen music even in the bathroom.
      c) I don't see any significant benefit of using the system.
      d) I will be confused when I cannot use it properly and this will make me nervous.
      e) I hate new technology; can't we just use a radio?

    12. How much would you be willing to pay for the system if it were commercially available? Indicate the maximum amount you would be willing to pay.

      a) $200-500
      b) $500-1000
      c) $1000-2000
      d) $2000-5000
      e) Other. Please specify:

    13. Do you like the idea of all your music being stored on a remote machine without you having access to your physical CDs?

      a) Yes, I can search a song faster than I was able to before.
      b) Yes, now I don't have to physically store hundreds of CDs in the house.
      c) I'm not sure.
      d) No, it is a little confusing.
      e) No, give me my CDs back.

    14. Which mode of operation do you prefer:

      a) voice
      b) touch screen
      Please explain:






    15. How important is it for you to be able to give very loose or vague voice commands vs. very specific commands? For example, would you like to be able to say "Play that song ... well you know it...the one that I heard last night" or you would prefer to be explicit "Play The yellow submarine by the Beetles"?

      a) Very important, I would not use it otherwise.
      b) It is good to have loose commands but this is not that important for me.
      c) I'm not sure.
      d) Having only a limited vocabulary certainly speeds up the interaction.
      e) I don't trust the voice interface; the touch screen is the only thing that I use.

    16. Which feature of the system did you like the most?







    17. Which feature did you like the least?








    18. Is there anything that was not mentioned above but you think might be important?







    Appendix

    Further readings and Miscellaneous Links


    Follow Me Home Page
    Last Modified 11/10/97 -- Jason Elliott