Initial Prototyping

This week we brainstormed and prototyped potential solutions to suit needs related to travelling. Below are some video prototypes of our favorite ideas. These were prototyped using 1) a “wizard” to manipulate Google Earth the traditional way (i.e. keyboard) in response to a user’ actions and 2) a projector to project the Google Earth view onto a large display.

Prototype #1: Visiting a Friend

Problem: It is hard for users to feel connected to other people living in other countries due to the geographical and time constraints.

Proposed solution: Virtual “windows” that connect to other places, such as another country or a long-distance friend’s home. Through these windows, users can observe and interact with other people using gestures. This particular prototype imitates a “space” in which users can interact with one another. The space looks like a home, in which the photos on the wall are actually photos of the homeowner and the visiting user. There is also a phone, which a visiting user can use to make a video conference call.

In this video, Hyunggu visits his friend in Japan. Starting from the global view, he navigates to his friend’s home. Once there, he knocks on the “door” to enter his friend’s “home.” He sees three pictures on the wall and selects one of them to examine in more detail. These pictures are of Hyunggu and his friend. Then, Hyunggu points to the phone on the table. This initiates video call with his friend, in Japan.

Issues: Users have trouble finding locations on the globe and maps (geographically-challenged) . Providing a list of locations to choose from would have been helpful. These locations could be sorted by user preference and/or recent history. Furthermore, special gestures for browsing photos and making video calls rather than direct pointing are needed to create a less awkward interface experience.

Technical Feasibility: There are videoconferencing APIs that allow calling from within the WPF framework. Google Earth offers an Javascript API that can be controlled from the WPF framework. Bing Maps 3D offers a deprecated SDK that can be embedded in WPF applications.

Prototype #2: Planning a Trip

Problem: Planning a trip is boring and time-intensive. Also, it is difficult to have a solid idea of what a place is like solely from textual descriptions and photos online.

Proposed solution: An interactive hotel booking experience in which users can save time and gain a realistic sense of what their travel experience will be like.

In this video, Tatiana acts out a simple travel planning experience. We wanted to see if planning a trip by using gestures could be more efficient and valuable than traditional travel planning methods. First, she navigates to her travel destination starting from the globe view. Once she zooms in far enough, the map view is scattered with icons denoting hotels. She selects her desired price range. She looks at a few possible hotels. The idea here was to eventually enable the user to take a virtual tour of the hotel.

Issues: Manipulating the virtual globe: hands cannot cross each other; moving hands back to starting position also rotates the globe back. It was also hard to find locations on the map. It would’ve been helpful to have labels or an additional text layer in order to facilitate the exploration process

Technical Feasibility: Google Earth’s Javascript API allows for the embedding of layers to hold hotel information. Google Earth’s API can be hacked to be used in WPF applications. Bing Maps offers a SDK that can be used in WPF applications.

Prototype #3: Virtual Tour Guiding

Problem: Being in different locations makes it hard for one person to describe what it’s like living in one place to another person living somewhere else.

Proposed solution: Use projected Google Maps Street View to walk through a place together, one person acting as a tour guide while another follows along remotely.

This video prototypes Daniel giving a tour of his home in Singapore. We wanted to see whether “walking” through a projected street view felt as natural as walking and talking in a real setting and whether or not viewing the tour was believable.

Issues: There are a lot of chances for “false positives”, as users tend to walk closer to the screen and wave their hands about. Needs a very clear method of controlling interface (walking around, “moving” head) while allowing the user to naturally gesture at the screen.

Users also (naturally) wanted to “point” at things on the screen. A cursor (or some other form) may be useful to highlight particular items on the screen.

Technical Feasibility: Google Maps offers a Javascript API, while Bing Maps Streetside offers a SDK. Highly possible.

Gesture Brainstorming: Zooming Techniques

Problem: A wall-sized display aims to create a sense of immersion. How might we enable someone to feel as though they are manipulating 3D space without using mouse/keyboard input?


This prototype features Daniel and Daniel trying out various gestures for zooming in/out while another person acted as a wizard controlling the Google Earth navigation. The role was to enable zooming in and out on a bird’s eye view map. The main focus was to test out how natural the gestures felt.
Gestures:

  • lean forward/backward: leaning back was awkward, uncomfortabe, and limiting; offered no sense of scale
  • step forward/backward: easy, not completely intuitive
  • frame corner manipulation: feels comfortable, defined bounds provides a better sense of scale
  • mixed – lean forward, step backward: confusing because they aren’t direct opposites
  • grabbing, releasing: clutch = grabbing, felt good, track elbow and hand
  • false positive: does dropping arm or unintentially pulling hand back fire gesture recognizer?
  • inertia – throwing feels good

Gesture Feasibility:
one foot forward/back

  • for walking
  • easy to detect with kinect and calculate difference
  • difficult to distinguish “forward” vs “back” without restricting control to one foot
  • false positives: user stepping around in place while talking could unintentionally fire gesture recognizer

hand in “hot zones” on either side of screen

  • to turn left/right
  • easy to detect because it’s a hit test

Additional Prototypes :

“Google Earth” as an interactive museum

Users navigate through an information-rich Google Earth. It is augmented with metadata such as Wikipedia articles or YouTube videos relevant to the location.

Thematic Restaurants

Eat authentic food in a restaurant, while feeling like you’re actually there. This is achieved through a wall display that acts like a “window” in a train booth. Landscapes are displayed. Other features might include a live video feed of a chef preparing your food. Instead of having to look at a physical menu, the users could order from the display.

A Day in the Life of…

Follow someone around for a day to see what their life is like. This has potential for cross-cultural exchange. A user could “participate” along with whoever they are shadowing. This could also be a way to practice language conversation or make friends.