Occupancy detection

["Mark Harrison" <Mark.Harrison@xxxxxxx>]

Hmmm... I've been thinking more about this.

At the moment, there's only one room where my "occupancy
detection"
actually works reliably. (Ie both always detects when I come in, and
doesn't give the "false exit".)

This room is my dressing room, and has the key characteristic that, when
I'm in there, I'm moving about. As such, a single PIR, cunningly
positioned, does the trick, together with an HV timeout of about 90
seconds.

This was all sparked from some things that Kieran Broadfoot said at one
of the previous London meets - in fact, the meet in which I uttered the
famous line about an Asian domestic servant :-)

Kieran's ideas have worked out very well, and I'm really finding that
the house WORKS! Remote control is nice, but this is WAY more useful
when it comes to human-house interaction.


In my study, however, I don't move about enough, which got me to
wondering about pressure pads. I have two chairs in the study, both of a
contstant weight.

Is it possible to get a pressure pad with a sort of "trigger
weight"
setting, that could go under the chair. Then EITHER coming into the room
(triggering the PIR) or sitting on the chair would give a "Room is now
occupied" event. But a "Room is now empty" event would only
be triggered
once BOTH a "PIR no longer detects anyone" and a "pressure
pad no longer
reports weight" events had happened.)


In Mary's bathroom it's even worse. After being cast into the dark twice
(once while on the loo, once while in the bath), she has taken down, and
hidden the PIR. With the bath, I can imagine that a pressure-pad
solution would work (ie if the bath is full of water, then the room is
deemed to be occupied, and don't worry about distinguishing between
"full of water" and "full of person".) On the loo, a
"break beam" type
of thing could tell if someone were sitting on it.


Any one got any thoughts about this.


Mark Harrison
Head of Systems, eKingfisher