The user waves from left to right. Our original design included the logo and at the bottom a hand with a trajectory. Users had no problem understanding what to do to continue. This is an indication of a good design for the first window. We later decided to add an animation to it for aesthetic purposes.

The user swipes to browse through color scheme and holds for 3 seconds to select the color. Our original design was mood based and we used pictorial themes and one song to display these moods. The drawback from this design is that we realized that people may not want to be restricted to four distinct themes. Using a color scheme enables users to pick an “in-between” mood which out feeling like it has to be one or the other. Along with the color scheme, the user will be able to select a song. We were debating whether it was worth it to have a full library or music or just have a few selections of music from different ranges.

Users are presented with the choice of reading news snippets or watching video clips. Based on the Wizard of Oz testing we concluded that people want an interface that is easy to control and digest. We tried to limit the amount of content and laying that a user has to go through in order to digest some kind of information. We wanted the UI to be accessible but at the same time satisfying. The design we decided to go with:

• The user is presented with the option of browsing through news article snippets or video clips
• The user is presented with different categories: different subjects for the news snippets or a range of videos.
• After selecting a category, the user is presented with either a range of news snippets or the video where they can start/stop it. We wanted to control the amount of layered content which is why the news snippets are the last layer or reading content.

Volume Control: Users are able to control the volume of the music and video through some kind of a dial that appears when the volume button is selected Color/brightness Control: Users are able to change the color and the brightness of the lights. We plan on integrating Phillips Hues lights

The start screen is the welcome screen for user in the bathtub. User can wave to enter the application while in bathtub.

After the start screen, users are brought to a customization page for choosing the environmental factors. We represent our themes in terms of colors - different color means a set of background+lighting+music. We picked 6 different color combinations, each including a large box - the background color, and a small box - the foreground color for text. Since currently we let user to pinpoint the color theme they want(by holding it for 3 second), we thought it would be easier for them to pinpoint larger target. Thus we splitted the six themes by showing three at a time, and users can change different color groups by swiping left and right. After that user is in the home page, the customization page will stay as a layover so that user can always come back to it later.

The Home page includes the “News” and “Videos” selection for user to choose. From here user can enter the category he/she is interested in. Also the “Color” and “Sound” customization choice stays as layover on the page for user to change later.

• 1 Start Page

• 2 Choosing Theme

• 3 Home Page

Other: Calibration Page - We figured out that we have to calibrate the user's position to maximize the reach on the screen with comfortable gestures. Therefore, we added a calibration page which quickly captures users' position and calculate the bound of reach.

• Calibration page

• We are currently using C# with support of an external Kinect library. The library is using DependencyProperty class to dynamically change the value of the attribute of the object. We feel that this class is very useful and important for us to refactor and structure our program, therefore, we aim to understand the usage of it and improve the computational efficiency of our code.
• We use C# with Visual Studio to program the functionality and Expression Blend to design our UI. These programs are Windows specific and have a steep learning curve. We learnt that we should be well planned ahead of time to have them installed and get ready to deploy prototype.
• We at first plan to implement lighting control with Philips Hue. Philips Hue uses json object and specific protocol to update the color change. In addition, the fact that we have chosen C# make it more difficult to work with the APIs of Philips Hue which are mostly written in js.
• There are still calibration issues when the users hover their hands over to select an object on the screen. The showing up of the hand location can be shaky and this requires better calibration algorithm.
• We started to think about the configuration of the application, such as where to put the screen, the kinect, and projector (if there is any). We realized that the set up distance in a normal bathroom may not be long enough for kinect to work perfectly.
• There are screen resolution dependence issue that we are aware of too. We may want to decide the screen resolution to decide for in order to avoid formatting issues during the final demo.

During our design and consistently talking to different people, we found out some issues with our design:

1. the app is trying to achieving many things - adjusting lighting, color, songs, reading news, watching videos, which are not only a lot to implement, but also weakened our main goal - when it’s doing so many things at the same time, what is the one thing that will fulfill the needs of our users - which is to relax? We will have to rethink the problem and come up with a clearer goal.
2. interface-wise, there are many different layers(e.g. main page -> news category -> browsing news -> individual article), which is very tiring for users in the bathtub(and might not be interesting to switch back and force). So we need to come up with a cleaner interface with fewer layers where users can interact with the system in a more natural way.