[extropy-chat] Worldwide SOS system

Mike Lorrey mlorrey at yahoo.com
Thu Oct 6 02:13:19 UTC 2005

--- Dan Clemmensen <dgc at cox.net> wrote:

> Emlyn proposed a system with the following characteristics:
> "transmitter" fits in a one-gallon container and has a very long
> shelf life. Receivers are in orbit.
> Rugged, robust, dead simple.
> I think we need to add: satellites must be cheap and offer coverage
> in a short time, preferably less than two hours.
> Sorry, Emlyn, but radio can do all of this for a lot less money than
> satellite imagery.
> Use a highly simplified radio that is designed to broadcast at a low
> bit rate in a narrow bandwidth. This yields a high Eb/No (energy per
> bit) even with low power. The assumption is that there are only a few
> transmitters at any given time. in any particular part of the world,

True, but this would need to be integrated into items with utility and
ubiquity outside of the subset of people who actually prepare for
emergencies. Ubiquity is the problem: having the item on hand when you
most need it. The problem with most people who are lost is that they
are unprepared for their ordeals, even those who intentionally go out
in the wild like campers/hikers, etc. Building the capacity you
describe into an item unprepared people are typically likely to have on
hand anyways is the product design psychology we need to approach it

Cellphones, for example, are radio transmitters that are ubiquitous
(for example, Nigeria is the highest growth market for cell right now).
Having a subcircuit designed into cellphones that has an emergency mode
like you describe if it can't lock on a cell tower wouldn't be hard.

As for power, most cellphones fold these days. Given the crank
generator technology, it shouldn't be hard to build such a small
generator into the fold hinge of cellphones, so even if you are lost
and with a dead cellphone battery, given enough folding over and over,
you could crank up enough energy to broadcast a mayday.

> You can get arbitrarily long shelf life for a simple battery. For 
> example. use a lead-acid battery, but keep the acid in a separate
> sealed container until you need to use the radio. You can get a lot
> of energy in a half-gallon container. The radio would be detectable
> from simple satellites for at least two days after you pull the tab.
> The satellite would notify the emergency response team, which would
> fly over the area to localize the transmitter. No need for GPS, cell
> phone, or even primitive AM radio, I think the emergency transmitter
> including container, battery, and acid, could cost less than $20.00,
> Add GPS, FM receiver, AM receiver for less than an additional $30.00,
> and a transponder system (to speed up localization after the response
> team gets to the area) for an additional $10,

The satellite needs to know a lot more than the general area the
transponder is in. Even if its smart enough to sense the doppler shift
in the transponder signal, it is only going to find a point of
strongest signal that is on a line somewhat perpendicular to the
orbital track of the satellite (depending on the orientation of the
transponders monpole antenna). It will take multiple passes over the
area to get any kind of a fix that is less than 100 km x 100 km. A
child can be lost in a 2 km x 2km area long enough to die of exposure
before anyone will find them. GPS is already built into all the newest
cellphones, so take advantage of it, it doesn't take too many bits to
transmit a GPS position at a low bit rate with minimal power draw.

> Fifty satellites at $100,000 each would give great coverage and would
> be  cheaper than
> a one satellite that is capable of the imagery your system requires.

Well, not necessarily. Doing it in a 1 kg, 100cm cube package would be
tough. I guarantee I could do it in a 3 kg, 100x100x300cm package, and
for a lot less than $100k, of course I happen to have a pappy with
experience designing spysat cameras, but that's just me.

First thing I'd do is use more than one CCD. I'd build a two by three
CCD sensor grid, using the new 8.4 megapixel sensors, which are
1"x1.65" in size. This would give me an effective image size of 7680 x
6480. When imaging the surface to a 1 meter resolution, which is pretty
common for commercial imaging sats today, capable of reading five meter
tall letters, this means imaging a 7.6 x 6.6 km area at a time. The
imager would require a 30x Dobsonian scope to zoom into that resolution
level. Challenging, but not hard.

The real limitation is on how fast the CCD can take pictures. Most
consumer digital cameras require at least a second to store an image at
full resolution. Faster flash ram is needed, and a faster bus speed.
Orbiting at 200 km in a polar orbit at 7.78 km/sec, you have a 40008 km
polar circumference of the earth and a 5303.8 second orbit time, you'd
need to take a new image every 0.97 seconds to have a continuous
imagery path throughout the orbit. As the earth rotates beneath the
satellite, you can image a swath of the earth 158.4 km wide at the
equator and 0 km wide at the poles each day with one satellite. This
means you need about 255 satellites to image every spot on earth each
day at least once. Except that half of this area is in darkness when it
is passed over, so you actually need 510 satellites.

This is the main problem with this concept at this point in time: not
enough resolution available for each satellite to image more than a
small area at one time. On the plus side, the closer you are to the
poles, the more frequently you get imaged due to overlap. 

Getting a realistic constellation of 25-50 satellites will require an
improvement in CCD resolution of 3-4 generations, a quadrupling in
flash ram speed. So this is a 2012-2014 time frame.

Of course, if you don't need to resolve every point on earth once a
day, it gets easier. If you don't mind once a week imaging at the
equator, 72 satellites could do the job with current technology. It
would give increasingly greater coverage in the temperate, subpolar and
polar regions, which, given the risks of exposure to cold with time,
the attention needs to be focused. 

Given all this, it is apparent that the radio option is the better one
for the time being for the most people. They can sense transponders
over a wider swath at a given time, and don't require a lot of OCR
processing, just a bit of doppler shift and signal strength analysis,
and it also makes the job easier for people lost at sea who can't
maintain large SOS letters on the open water.

The imaging option does deal with those stuck without functioning
technology that is up to date, though. Given how easy it is to destroy
a cellphone with a splash of water, the low tech imaging option should
not be dismissed out of hand.

Mike Lorrey
Vice-Chair, 2nd District, Libertarian Party of NH
Founder, Constitution Park Foundation:
Personal/political blog: http://intlib.blogspot.com

Yahoo! for Good 
Donate to the Hurricane Katrina relief effort. 

More information about the extropy-chat mailing list