Thoughts on home sensors for the deaf
Ken
I've been thinking about the project you mentioned yesterday. It seems
like a valuable, cool idea. At the same time consumer products are a
highly competitive area.
I notice a company called Silent Call (silentcall.com) that has
wireless products for sensing and notifying. Given that there is at
least one company doing something similar to what you are thinking, I
would think about how to do it differently.
Sensing: I wonder if it would be possible to install a generic
listening device in important rooms, and then interpret the nature of
the sound via signal processing? What is the accuracy of something
like this? Could the system be trained on certain sounds? I see the
Silent Call people have a generic monitor.
Another option is to look down the road to Zigbee or ZWave home
automation. Are there any prototypes or specifications for interfacing
to appliances?
Internal communication: Besides Zigbee, Wifi seems like an easy way to
get a signal to an actuator.
That's all I can think of for now.
Good luck,
Ken
Ken
I know where you're coming from. It took me most of last summer to
talk myself out of the original idea I had for my project. :-)
At the same time, I wouldn't say that you're trying to fit the network
to your problem. As I understand it, the ZigBee project is highly
focused on home automation. It will take a while for it to become
popular, if it ever does, but it ought to be mature enough now for
this sort of research.
Here is a recent article from IEEE Spectrum: http://
www.spectrum.ieee.org/oct06/4666. One of the authors is local:
JON ADAMS is director of radio technology at Freescale Semiconductor
in Tempe, Ariz.
There's no harm in writing to him about your idea. Who knows what
you'll find out. These companies are always looking for smart CS
students, too.
Ken
Corby wrote:
> Actually I had almost decided to give up on this idea yesterday.
>
> I was thinking about how to implement something like what I had planned and the best solution seemed to be to have sound sensors that are sensitive to high pitch or high amplitude sound, and place them near devices that make those sounds. Then have it transmit to a central hub that broadcasts it using RF to whatever pager/watch or alert device you have, which then has the codes pre-stored to display the message.
>
> The problem when I arrived at that idea is that it was no longer an interesting networking problem.
>
> I was aware of the systems Silent Call (and Clarity Alertmaster, Sonic Alert, and Ultratec Simplicity) make, which is what I was referring to yesterday from research (a good page to see the things that are available is deafbuy.com <http://deafbuy.com>). They still seem like pretty primitive systems to me. So I was thinking how to make a big improvement on these existing systems.
>
> After awhile though, and considering how X10 and these existing systems work, I was starting to notice I was trying to fit the idea to the type of network I wanted to study, instead of choosing the best network for the problem.
>
> So right now I'm thinking of other ideas, but I'm also going to go back and think more about the initial problem with the ideas you've brought up. Hopefully I can come up with something that's useful *and* an interesting networking problem.
>
> -Corby
>
>
>
> On 1/25/07, *Sandeep Gupta* <Sandee...@asu.edu <mailto:Sandee...@asu.edu>> wrote:
>
> Corby:
>
> How about also thinking about location-awareness (context-awareness) and
> trying to reach a person based on his/her current location/status (e.g.
> if a person is taking a shower then flash a light (since I would
> presume
> people wouldn't like to take shower while wearing a watch).
>
> Sandeep
Ken
index.php?page=home.
The publications page references a paper describing a routing protocol
for it:
http://prisms.cs.umass.edu/brian/pubs/burgess.infocom2006.pdf
Ken
Corby wrote:
> Doing the whole location awareness thing really gets into the smart home applications that I've noticed other people have done, there are lots of papers on these things I came across when trying to find information for the deaf system. It also really gets away from my initial idea of something that could actually really be a real product and soon... something affordable and practical, that's not a pain to install (no ripping up the walls or carpets).
>
> As for other research ideas. I was struck by the idea we covered last week of using a bus to pick up news or some sort of data like that, and transporting that data to a rural location where it would automatically download to an access point there. Sort of a way to get around bad network connectivity. However I don't think that idea in itself is practical because rural locations can use 56k modem, satellite or cellular service (like Sprint Broadband).
>
> But generalizing that idea of using some physically moving object as part of the network intrigues me, and I was wondering if you knew of any papers on such a topic?
>
> I think it might have applications for maybe a way to covertly transmit sensitive data in an insecure network environment... maybe something CIA related. Although my mind is wandering thinking of other applications.
>
> -Corby
On Feb 12, 12:34 pm, "Ken" <r...@deru.com> wrote:
> Corby,
>
> I know where you're coming from. It took me most of last summer to
> talk myself out of the original idea I had for my project. :-)
>
> At the same time, I wouldn't say that you're trying to fit the network
> to your problem. As I understand it, the ZigBee project is highly
> focused on home automation. It will take a while for it to become
> popular, if it ever does, but it ought to be mature enough now for
> this sort of research.
>
> Here is a recent article from IEEE Spectrum: http://www.spectrum.ieee.org/oct06/4666. One of the authors is local:
>
> JON ADAMS is director of radio technology at Freescale Semiconductor
> in Tempe, Ariz.
>
> There's no harm in writing to him about your idea. Who knows what
> you'll find out. These companies are always looking for smart CS
> students, too.
>
> Ken
>
> Corby wrote:
> > Actually I had almost decided to give up on this idea yesterday.
>
> > I was thinking about how to implement something like what I had planned and the best solution seemed to be to have sound sensors that are sensitive to high pitch or high amplitude sound, and place them near devices that make those sounds. Then have it transmit to a central hub that broadcasts it using RF to whatever pager/watch or alert device you have, which then has the codes pre-stored to display the message.
>
> > The problem when I arrived at that idea is that it was no longer an interesting networking problem.
>
> > I was aware of the systems Silent Call (and Clarity Alertmaster, Sonic Alert, and Ultratec Simplicity) make, which is what I was referring to yesterday from research (a good page to see the things that are available is deafbuy.com <http://deafbuy.com>). They still seem like pretty primitive systems to me. So I was thinking how to make a big improvement on these existing systems.
>
> > After awhile though, and considering how X10 and these existing systems work, I was starting to notice I was trying to fit the idea to the type of network I wanted to study, instead of choosing the best network for the problem.
>
> > So right now I'm thinking of other ideas, but I'm also going to go back and think more about the initial problem with the ideas you've brought up. Hopefully I can come up with something that's useful *and* an interesting networking problem.
>
> > -Corby
>