Re: proximity/occupancy sensing and PC integration?

[Dennis Slater]

Somewhat out of the box: I think that a promising method to detect a
person's presence or whether someone entered/left a room would be using
odor classification and location technology. I believe each of us has a
distinct odor. If a person's odor could be detected and matched with a
database of odor characteristics a person's presence could be recorded.
Or if the base odor characteristics of a room were changed by the
arrival of a new source of odors that event could be recorded.

I have no idea if odor detection technology has reached that state of
sophistication or not. I believe Cal Tech is working on a 'silicon
nose' that will result in "the construction of a single chip
neuromorphic electronic silicon nose capable of odor classification and
location". Without future Googling I do not know where they are on this
project.

[wkearney99]

> I have no idea if odor detection technology has reached that state of
> sophistication or not. I believe Cal Tech is working on a 'silicon
> nose' that will result in "the construction of a single chip
> neuromorphic electronic silicon nose capable of odor classification and
> location".

It's one thing to have a detector in a controlled environment. Passing
samples to it with a certain degree of control over what's presented. But
trying to do the same thing in a household environment seems like a much
bigger problem. Mainly due to the uncontrolled nature of what samples might
be presented. A dinner burning on the stove or, worse yet, household
cleaning solvents would pose considerable challenges to such a unit. Not
just from a detection standpoint, which would be bad, but from a clean-up
perspective. Thus I don't forsee odor detector being practical in a
household environment, at least not for fine-grained detection of user
presense. For fire or gas warnings (natural or carbon monoxide) they're
certainly useful. But being able to tell if it's you, me, the dog or the
cat doesn't seem terrible likely.