TUI Fall 2013

Think Inside the Box

And we're done! The feedback received after the presentation was great, and extremely encouraging. CS320 may be over, but the MagicBox team will be back in mid-January to work on the rest of the prototypes for the TEI conference in February.

MagicBox was a collaborative effort that was possible due to the amazing people involved in it. It is hard to say who did what, as all three of us were very involved in every aspect of the project. It wasn't until we started box 2 that we divided the work. I ended up focusing on box 1, as we had to revisit the design to include the speakers and photocell sensor.

While we were still trying to figure out how to use the IMU, we decided to start box 2. When we designed box 1, we didn't consider modularity, but since we got accepted to TEI in Munich, our professor told us we should think about how we're going to transport our TUI.

Box 2 uses 2 foam boards that can be placed in any box large enough. In addition, this prototype is much more organic, as it doesn't use an Arduino or any fancy sensors. Instead, we used felt, tulle and conductive thread to make the soft switches. The LED lights turn on when the switch is pressed.

After doing some research, we decided to test an IMU instead of a gyro + accelerometer combination for the spinning interaction in box 1. Bad idea. We ordered the DIYDrones ArduIMU+ V3 IR proximity sensors, and 0.5W (8 ohm) speakers.

Unfortunately, we couldn't get the IMU to work, and later on realized that we could have probably just used a gyro. We looked around the lab once more, and found a photocell sensor. We realized we could use it to control the sound, based on the intensity of the light. By placing it at the bottom of the box, every time the user's head moves, the tone of the song changes. We attached a second breadboard, as we needed to connect the 2 speakers and photocell sensor to the Arduino.

We started working on scenario  (medium box, over user's head)

Since we're constrained by the resources we have in the lab, we chose to use a button on the side of the box to simulate when the user puts the box over their head. When the user presses the button, a LED light turns on. This was quite straightforward and easy to implement, as the Arduino software comes with sample codes.

However, our biggest challenge is how to make the wires stay connected to the Arduino as the user turns the box upside down.

At first we tried soldering the wires directly on the breadboard, but we ended up just melting the breadboard and couldn't get the melted lead inside the holes fast enough.

We came up with 5 major user scenarios, each one of them designed for a specific type of interaction. For each scenario, we want to use different types of boxes. For example, scenario one would require a medium sized box that the user can put over their head, and manipulate easily with their hands. Another scenario requires a larger box, so that users can trigger the outcome as they crawl through it.

We have decided to use Arduino microcontrollers, as we want to be able to use a wide variety of sensors. User scenario 5 is the only one that would require a different type of technology, since the interface we want to implement would be easier to do with RFID tags than with Arduinos.

The pictures below show all 5 scenarios:

Our proposal tackles the design prompt related to magical experiences. We aim to create a magical experience by allowing children to play in an environment with no set rules and previous knowledge of the context. We plan on using cardboard boxes and sensors, which will trigger specific reactions as children interact with the environment. We chose this solution because we want to explore the idea of intuitive interfaces, in particular among children, who do not have a clear understanding of their surroundings yet. Because often times children are more fascinated with the packaging of their new toys, we argue that this is due to the lack of constraints they encounter when they play with boxes. Most toys for children have a specific goal (complete a puzzle, play a tune, etc), or they try to mimic an experience from the real world and revolve around role playing. On the other hand, our solution attempts to develop their imagination by not having any rules. We also think that this project will encourage social interaction, as the environment will be better used if more than one user interacts with it.

PICO (Physical Intervention in Computational Optimization) is an interface that utilizes mechanical constraints and abstract computational power. Mechanical processes have limited resources to accomplish computational tasks; in contrast, modern computers are able to perform more complex operations. However, mechanical systems are better designed for collaborative frameworks, as users have physical objects that are easier to manipulate when working with other individuals. With that in mind, PICO was developed to test an application for mobile networks.

 From a Reality-Based Interaction (RBI) framework, all four themes are present in this interface. The most obvious one is the user’s intuitive knowledge of naïve physics, which is observed when users employ physical constraints to manipulate the physical world. In addition, the user is not constrained by the objects the interface offers, and in fact, they can use almost anything to interact with the interface. For example, the user can control the friction by adding weight on the puck or spreading soap on the tabletop. Moreover, the user can manipulate the distance between two pucks by using collars and rubber bands. The interface relies on people’s notion of the world as they know it, and their literal interpretation of their surroundings in order to be able to apply physical limitations to the system. In addition, in terms of body awareness and skills, the user is able to interact with the system by using both hands, and use their dexterity to manipulate the pucks. The user also utilizes their environment awareness and skills; they have a sense of physical space, as they are aware that the pucks can be moved on the surface, and can also physically move their own bodies around the tabletop and control the pucks from different angles. Furthermore, from a social awareness and skills perspective, the interface encourages social interaction by not having a single input, which allows multiple people to be able to move the pucks.

 PICO presents some tradeoffs when used to simulate a cellular network. In particular there is a reality vs. practicality tradeoff, in favor of practicality. The mechanical constraints do not exactly model the real world, and for example, height of buildings intercepting the tower’s signal is hard to measure using PICO. Moreover, there is a high level of abstraction because PICO was designed with the intention of using it for multiple applications, consequently, there is also a reality vs. expressive power tradeoff, as the pucks can represent almost anything depending on the software configuration.