I would like information about any smoke detectors which make use of
the following technology:

http://www.freepatentsonline.com/4206456.html
http://patft.uspto.gov/netacgi/nph-Parser?Sect1=PTO1&Sect2=HITOFF&d=PALL&p=1&u=/netahtml/srchnum.htm&r=1&f=G&l=50&s1=4,112,310.WKU.&OS=PN/4,112,310&RS=PN/4,112,310

Briefly, the smoke detector is light sensitive and becomes more
sensitive at night. This reduces nuisance alarms during the day yet is
more sensitive at night when most fatal house fires occur.

I've looked and haven't found any smoke detector which does this. This
seems like a good idea and shouldn't be expensive to implement. Why
wasn't this done, especially since the patent is now expired? Are
there economic or legal reasons this would not make a good product?

I have some money to invest and I would also like some advice as to
whether investing in developing a smoke detector as described above
would be a good idea. Would it be a successful product? How much money
would be required? How should I proceed given that I can't get a patent
for the idea?

Link to study conducted by NIST on Optical Fire Detector vs current cheaper tech.

http://smokealarm.nist.gov/

Honeywell makes one

http://www.ademco.com.au/Products_Fire_FDD_2151.asp

Search strategy : Optical smoke detector

I question how you would (or if you could) sell such a device to the
public. If the sensitivity during the daytime is attenuated, you run a
serious liability risk. Furthermore, winter is probably when most
house fires occur-- celebrating Christmas seems to be inviting fiery
Death, according to firemen-- and it is dark most of the time then.
You make breakfast while it is dark, you make dinner while it is dark,
you are gonna have false alarms while it is dark. Finally, if the
thing is photosensitive, turning on a light near it makes it think it
is daytime and it will operate at reduced sensitivity.

But an intriguing idea is one that requires agreement from multiple (2
or 3, not too far apart, maybe 6 feet) sensor locations before
sounding an alarm. Or a detector that increases sensitivity during,
say, midnight to 6AM. Heck, you could jury-rig up such a thing
easily-- have added detectors connected to a timed power source that
only comes on at designated hours while one or two are always on.

To be safest you really need an alarm to be stupid.

Isn't the Honeywell photo-optical smoke alarm the same as the old
photocell alarms? Those used to be set off by not only cooking smoke
but shower steam as well. The "sniffer" types were not fooled by steam
but were more sensitive to cooking smoke.

> Honeywell makes one
>
> http://www.ademco.com.au/Products_Fire_FDD_2151.asp

This is not what I was asking about.  There are two common technologies
used for smoke detectors, photoelectric and ionization:

http://chemistry.about.com/cs/howthingswork/a/aa071401a.htm

What I am asking about is a smoke detector which uses either of these
technologies in combination with a photo-electric sensor to adjust the
sensitivity of the detector, making it more sensitive at night, where a
smoke signal is more likely to be caused by a real fire.

> I question how you would (or if you could) sell such a device to the
> public. If the sensitivity during the daytime is attenuated, you run a
> serious liability risk.

Yes, liability risk is an issue. However if a detector were less
sensitive during the day and more sensitive at night, there might be a
net positive benefit. There are more fatal house fires at night than
during the day.  I suspect one reason smoke detectors do not always
work properly is that people remove the battery because of too many
nuisance alarms. Less sensitivity during the day might help here.

> Furthermore, winter is probably when most house fires occur--
> celebrating Christmas seems to be inviting fiery Death, according to
> firemen-- and it is dark most of the time then. You make breakfast
> while it is dark, you make dinner while it is dark, you are gonna have
> false alarms while it is dark.

I'm suggesting the optical sensor would be sensitive to inside lighting
levels, not outside.  Inside the house when people are active there are
usually some lights on. Even when making breakfast early in the morning
there is usually more light in the kitchen than at 3am.

> Finally, if the thing is photosensitive, turning on a light near it
> makes it think it is daytime and it will operate at reduced
> sensitivity.

Yes, if someone leaves a light on all night then it wouldn't work.
Of course it's not perfect, but I think in most cases it would help.

> But an intriguing idea is one that requires agreement from multiple (2
> or 3, not too far apart, maybe 6 feet) sensor locations before sounding
> an alarm. Or a detector that increases sensitivity during, say,
> midnight to 6AM. Heck, you could jury-rig up such a thing easily-- have
> added detectors connected to a timed power source that only comes on at
> designated hours while one or two are always on.

Yes, this might be a good idea if all the detectors were not overly
sensitive. Having 2-3 detectors going off in a nuisance alarm would be
really annoying. It's also a bit more complicated and expensive to do
this.

> To be safest you really need an alarm to be stupid.

I disagree. I think a smart alarm which is able to sense when people
are home and if they are active would be better.

This is semi related.  See the following for a potential detector that
has patent protection that has not been commercialized:

http://answers.google.com/answers/threadview?id=736260

http://www.nextedison.com/