Project 1: Observation of a Location You're not actually going to implement the first project; you have free reign to propose anything you want. However, you will be building a physical model of the device you're proposing, or creating a rendering/animation/video if your proposal is not a physical device. You will also be expected to prove that any technology you propose exists, either by citing other working examples, or by describing in depth how the technology would be developed.
Week 1: Observe the location and activities
Choose a space in which there is a reasonable amount of activity. Why do people use the space, and what do they use it for? Observe the activities that happens there. Use pictures, videos, sound recordings, drawings, and any other form you need to in order to describe space and the activity. Ask people in the space about what they do there, what they like about the space, what they dislike, how well the space works for their purposes. Look for patterns in the activity. Identify how the space and the activity are linked. Decide how you can sense the action and create a reactive or interactive system to enhance or change the activity.
Week 2: Propose a prototype
Imagine a system to realize the change or enhancement you came up with. Find ways to use the space as it exists, rather than modifying it, unless it's absolutely necessary. Your goal should be to make a system that is as dependent on the location it's in, and uses it as well as, the activities you've observed.
As a passive example, you could sense how many people use each entrance to a given subway station, and how frequently. As a more active example, you might want to give people a surprise in their daily transit through a public square or lobby. In order to surprise them, first you'd have to observe and collect data on what the normal pattern of activity is, then make an output system that changes that pattern. Perhaps you want to make a sculpture that changes the patterns of pigeons in a public square (a la Ed Bringas). Or perhaps you just want to sense how long people use the bathroom, and tell them when they've been in there too long.
The activity sensed and the viewer observing the results can be two different people, species, etc. In other words, you don't have to close the interactive loop between the actor you're sensing and the viewer to whom you're presenting the results.