Course Announcement

MAS 961: Special Topics in Media Arts and Sciences
Developing Applications for Sensor Networks

Monday and Wednesday, 2:30-4:00 PM
E15-335

Summary:

This class provides the background and tools necessary for students to build and program embedded sensing devices that will both leverage and become a part of a broad distributed sensor network. Students are expected to complete a final project comprising the design, construction and demonstration of a novel sensor network application that builds upon a seed wireless sensing platform to be provided by the Responsive Environments Group. In addition to learning tools and techniques, this class will explore published literature concerning applications and user interfaces for sensor networks. Students are expected to participate in discussion of the assigned papers, as these will seed the development of their final project.

Prerequisites:

Permission of the instructor. In the first lecture, we will request students to submit a 1-paragraph justification for taking this class - as our hardware resources are restricted, we will select a limited number of candidates from this pool.

Instructor: Prof. Joseph Paradiso
TA: Josh Lifton, Responsive Environments Group
Units: 0-12-0

Background:

Many visions of ubiquitous/pervasive/invisible computing rely heavily on the assumption that people will one day inhabit a world dense with distributed networks of sensors and computational resources to process those sensor data. However, such distributed sensor networks suffer from a form of the chicken and egg problem: A great advantage of sensor networks over traditional sensing architectures is their flexibility and agility to perform multiple tasks at once and be reused in the future for tasks not considered at the time of deployment. Yet, no single sensor network task has presented itself as being compelling enough to warrant the considerable expenditure of designing and deploying a truly ubiquitous sensor network. In other words, sensor networks do not yet have a "killer application." This situation is both widely recognized and widely ignored by the sensor network research community. A practical way around this problem is to build up sensor networks incrementally and purposefully; incremental in the sense that nodes can be easily added over time to a small initial network and purposeful in the sense that every node in the network is designed for a real-world application and is of itself a useful artifact when considered alone outside the network. This class is the beginning of such a process. The sensor network built upon throughout this class is expected to continue as a viable research and application platform available to class participants, and possibly the entire laboratory, beyond the timeframe of the class itself.