Constellation Pose

Constellation Pose was a gesture-based Kinect installation that gives the audience the playful challenge of using their body to match different constellations and learning about the actual 3D shape of the constellations. 

 

 

Group Member 

Eojin Chae (Lead Digital Artist)

Laura (Writer)

Barry (Experience Researcher)

My Role 

Gesture Design 

Unity Development

Design Tools 

Kinect

Unity

Project Background

According to a recently conducted research of the Hall of the Universe in AMNH, compared to other exhibitions, the main screen in the hall is of least interest to the audiences, measured by average stay duration and pictures taken by audiences. As a result, the Science Visualization Group Emerging Media Lab decided to experiment with ways that could make it more playful and interactive. 

After the project was initiated, there were many discussion about the guidelines for design and development. And here were the basic guidelines that we established:

Contents

1. It should focus on being educational

2. It should fit into the rest of the exhibitions in the Hall of Universe

Interaction

1. It should be engaging and utilize the advantage of gesture interaction

2. It should try its best to be accessible to main audiences of the Museum, including kids, senior, people who speak different languages and people with disabilities

Gesture Iterations

Analyzing the first version

When I first joined this project, it just went through its first user testing with the public. So as someone who is really interested in the UX aspect of emerging technologies, I was lucky to have video documentation of what people were trying to do with the visualized universe.

By analyzing the video documentation, I realized that this gesture might not be the ideal gesture for zooming:

1. This gesture lacked a clear metaphor that fit into the context of space travel

2. This gesture did not best match the process of zooming

Zooming is a continuous and directional process. However, the first version of gesture required detecting change between two states, from arms open to arms closed.

In order to match the continuous process of “zooming”, users had to do this gesture repeatedly. Since the change between two states is highly reversible, during user testing, we realized that this gesture might trap users between two states.