On developing ubicomp applications

(eating your own ubi-dog-food is not enough)

Some three and a half years ago I was negotiating which research group to join, and got sold the idea of context-awareness. Somehow along the years it has got intrinsincly linked with ubiquitous computing: the computationalization of the everyday. One reason why context-awareness is so tightly connected to ubicomp for us is probably that we lack the people and know-how to do low-level context-awareness: signal processing, activitiy recognition, sensor development–instead we have to use high-level sensors like location and application use, which means that the activities last for a long time.

So anyway at some point we noticed that we were building and testing applications to be used throughout people’s life: in all places, situations and at all times. The thing is, to eat your own ubicomp food not only do you have to accept these applications into your everyday life, but also the development of them. You have to use your own applications, but to have a chance of improving their faults or capitalizing on their strengths you have to also think about those in your everyday life. Every time you notice a bug or a chance for a feature you have to stop and analyze it in that situation long enough to remember what it was about.

And so I am back in the situation I came to the university to avoid: having my work life intrude on my personal life beyond my control. Well, at least I did it to myself this time.

Wait, that’s not an upside!

This entry was posted in CS. Bookmark the permalink.

Comments are closed.