[ExI] old software fun

[Mike Dougherty]

On Tue, Jun 2, 2015 at 1:23 AM spike <spike66 at att.net> wrote:

> It has me thinking.  Back in the day, I thought how cool it would be to
> have
> a data logger at home, all the fun things I could measure and log.  Now I
> can have a handful of them.  So now my challenge is to get one which logs
> the old fashioned 0 to 5 volts, then I can set up my instrumentation to log
> whatever I want.
>
> I already have an application in mind: get a microphone, set up some op
> amps, filter it to center around 240 Hz or a bit higher.  A bee's wings
> sound to me like about third octave B.  (Hey, is that how bees got their
> name?  When they buzz around, it sounds like third B?)  Then we could rig
> it
> to trigger a data point whenever the microphone reached a threshold
> equivalent to a hovering bee 10 cm away.  Then if we could get those cheap
> enough, we could create a standard instrument for 20 bucks or so which
> would
> count bees.  That would be way easier and cheaper than anything else I have
> thought of.  Other bugs are different pitches, so we could band-pass filter
> the signal from about 230 to about 250 Hz.
>
> Alternative: just take data points every second with the microphone in the
> local bee playground, log the intensity of third B.
>
> Next I need to find a good cheap 0 to 5 volt data logger that plugs into a
> USB port.
>
> I can think of a buttload of cool applications around the home.
>
>
I wonder if video would be more useful data.  We're not very good at
processing sound (in general) compared with video.  No doubt you could
train an expert system ("deep learning" or whatever we call this flavor of
AI these days) how to detect bees from a video stream.  Instead of guessing
at the number of bees from a phased array of microphones for distances of
sound sources and cumulative intensity, it could count.  We might be able
to mine these streams for other insight we might not have otherwise had a
preconceived test.  ex:  it might be interesting to see what other insects
are detected when bees are either plentiful or scarce.  (or for a better
example, imagine something else we didn't think of then also remove it from
the set of all possible things that might be relevant but unknown.)

To further nerdify this setup we have to include some way to gamify the
data with crowdsource interaction.  I imagine some sort of "bee keeping"
analogy where the gamer is tasked with tracking/tagging the activity of
visible bees across the spectrum of howevermany camera feeds we have per
location.  Leaderboard plus achievement/badges should be enough competition
to keep eyes-on for at least a while.  There might exist opportunities to
monetize by selling virtual billboard space in a foraging field overlay:
advertisers pay a fixed price to get on the board, then pay additional for
each bee that cross their adspace with the premise that our beekeepers are
actively tracking the bees through those overlays.  "hive mind" + "mind
share" = "hive mind share"    I guess the next step would be AR/VR so
everyone can experience full immersion into both the bee problem and bee
solution.

ok, so there's some ideas... feel free to implement.  :p
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.extropy.org/pipermail/extropy-chat/attachments/20150602/5e922ae7/attachment.html>