Your Life On Disk

Photograph of Jason Brotherton

Jason Brotherton

The mantra of successful ubiquitous computing research has always included, "Build, Deploy, and Evaluate." The key to building these ubicomp applications is to know as soon as possible if your idea or design is flawed (fail fast), and if so, how best to fix it. This is achieved through rapid prototyping and evaluating the impact of adding the technology in real world settings.

This is all fine and good, but the truth is that there are few examples of good ubicomp research (based on this mantra) in the literature. I claim that the reason for this is the applications built need to provide the services of many different applications, which are often complex programs in their own right. Then, additional code has to be written to 'glue' these smaller applications together. In other words, the applications needed in ubicomp research are hard to build. (They are also hard to deploy and evaluate!)

In this talk, I will highlight two very different approaches to doing successful ubicomp research. One, the approach just described, is done in ivory towers with researchers. Another, more daring approach is to involve end users in the entire 'build, deploy, and evaluate' process.

Surprisingly, we believe this approach shortens the life cycle time and creates more usable ubicomp applications designed, conceived, and tested by users, but built by researchers. As an example, we will demonstrate this approach applied to the design of accessing your life on disk.
Tuesday 21st June 2005, 14:00
Robert Recorde Room
Department of Computer Science