Research Statement


In his seminal article The Computer for the 21st Century, Mark Weiser put forth a vision in which computers are ubiquitous, but so well integrated into human interactions that they go unnoticed or become invisible. By its very nature ubiquitous computing is interdisciplinary [Scholtz01], combining the fields of artificial intelligence, machine learning, sensor networks, distributed systems, embedded computing, human computer interaction, and even the arts [Abowd01].

It is now economically and technologically feasible to deploy and interconnect numerous sensors, actuators, and computing elements into everyday environments. However, how to assimilate the data from these sensors and make intelligent decisions based on that data remains as much an open question today as it was when Weiser first wrote his article. In other words, how to author interactive physical environments remains an open question.

Activity in physically interactive environments is inherently situated [Intille04]. Thus it may often be impossible to perform this research solely in the lab. For many applications work must be performed in situ. Artworks are deployed in museums, galleries, public spaces, and/or theaters. They are used by a combination of trained users (‘performers’) and an untrained public. Because of this, artworks must meet real world demands that cannot be tested in a laboratory or simulation. At the same time, artistic works usually allow a certain amount of control not possible in full ‘real world’ deployment. Thus artwork can provide a transitional venue for testing of technologies, more demanding then the lab yet not totally unconstrained.

My artwork is a test bed for advancing the fields of ubiquitous computing and human computer interaction. I seek to harness artificial intelligence approaches to author compelling interactive experiences. This is evident from my first collaborations, in which I developed a behavior based AI system for authoring interactive virtual-environments for the Bush Soul, to my current dissertation work, in which a distributed semi-supervised learning architecture is used to author interactive physical-environments.

The visual, performing, and interactive arts can serve as a catalyst for discovery in computer science. While artwork may server as a capable deployment domain for my research, it is equally true that the arts are the primary driver of my research. Artwork is a form of self-expression, a form that often necessitates innovation. My research seeks to address deficiencies in the research literature exposed by my participation in the arts. In my work, I strive to use advances in computing technology to expand what is possible in the media arts while, simultaneously, I use the arts as a driver to further developments in computer science


Eitan Mendelowitz