[aprssig] solutions for indoor tracking

[Wes Johnston]

In the past couple of days, a few suggestions have come forth.... This
is just me thinking aloud.

1)Create a fake GPS satellite to be mounted above each doorway that
streams out the GPS coordinates of the room you are in.  An IR receiever
could be added to a tiny track, pocket track or kenwood D7 for about $4
in parts.  The GPS sentence transmitter would consist of a $2 pic chip,
a $.25 IRLED and a $.30 555 timer.  At 4800 baud, the fake GPRMC
sentance could be sent about 7 times per second.  Anyone entering a room
would have this signal beamed down onto them and their tiny track or d7
would think it had just seen a valid GPS fix and would use that position
until another superceeded it.  This requires modifying your tracker so
it can read IR data, but once assembled is one unit.  This method will
make tracking lots of users a little expensive... about $100 a pop for a
pocket tracker with IR receiving capabilities, $160 if you want a pocket
tracker with a GPS.  You probably need to add $150 for a digipeater to
get on 144.39 to be placed somewhere in the building

2)Jim Lux suggested that we put the smart in the digipeater.  If a
digipeater sees a position that is invalid (ie GPS signal unavailable),
it would substitute a valid position.  This gets back to 1998 when
everyone did indoor BOXn-n tracking at Dayton.  You use a really deaf
radio like the MFJ data radio and make it even more deaf by clipping the
PIN diode.  It can then hear a D7 from 20 feet away at 5 watts.  Problem
is, today, we have pocket trackers that put out .25watts, and would be
heard only inches away.  But Jim's solution would not require the user
to carry any additional equipment.  A big plus.  Another problem is that
once an APRS client sees you with a valid position, and you later walk
into a building, it will continue to show you at the entrance to the
building until (in the case of aprs dos) for 80 minutes (hope I'm not
mis-representing this).  After that time is up, vicinity tracking takes
over and your invalid position data is ignored and you would then be
plotted near the digi that heard you... presumably the really deaf digi
in the middle of the room you are in.  This will probably cost $100 per
user and $150 per room.

3)Bob suggested that we use RFID badges.  They are cheap.... for
example, parallaxinc.com sells them for $2.50 each....and the retangle
ones look like name tags you'd clip on your lapel.  The reader is cheap
enough.... $40.  problem is it only works up to 4inches (100mm) and it
will get confused if there is more than one tag in range at a time.  The
cheapest long range reader I've seen is $300 in the form of a CF card. 
I must assume that it would not get confused in the presence of more
than one card at a time since it's range is so great.  Long range badges
can be had for probably $25 each.  I think this solution will be
prohibitively expensive on the building side, but the tags are cheap. 
If we were to use the $40 parallax reader, I think the think to do would
be to pin the RFID badge on everyone's right shoulder, and since we
(americans) tend to walk on the right side of a doorway, it's a good
chance your badge would come within 4 inches of the reader.  Or mount
the reader in the center of a door if you must open a door to enter a
room... spring loaded doors tend to get opened just enough for a person
to fit thru, forcing the reader to be near the badge.  We have to work
out a way to either program the badge serial number with a callsign, or
map serial numbers to callsigns on the client end.  Cost looks to be
around $3 per user, and either $45 or over $300 per room, plus about
$100 per building for a transmitter to get on 144.39.

4)Bob also suggested that we have IR badges that simply transmit our
callsign and that an IR receiver in the middle of the room (or ceiling
near the doorway??) would see your badge's callsign and translate that
to a position.  Something like a pocket tracker could be modified to
append a position to the end of a callsign heard on it's serial port. 
This is neat, and since I saw all those flashing callsign LED scrolling
name tags at Dayton this year, and they supposedly run all those LEDs
for 30 hours on a pair of watch batteries. It would seem that an IR
emitting badge would work for days if it only ran one IR led.  This is a
really cheap solution!  Especially if we find a way to do the
"digipeating" over twisted pair so that more than one IR receiver can
talk to a TNC at a time.  Cost is $5 per user, and $10 per room plus
some wire to get to a $100 transmitter in the building to get on 144.39

5)Brian suggested that we use 433mhz keyfob transmitters.  I presume
they would transmit kind of randomly like APRS trackers do now.  The
keyfob receiver would have the smarts to insert your position..  Keyfobs
that fit in your pocket go for $20 but they only send 5 bits of data,
and the receiver goes for $70.  433mhz transmitters go for $40... still
cheaper than a pocket track.  433mhz receivers are $60.  Also, you need
to add a $100 transmitter per building to get onto 144.39.

So we have 5 solutions.... 3 of them require you to carry a different
tracking device indoors and 2 of them will allow you to enter a building
and be tracked with the same equipment indoors and out.

Wow, this has been a great thread.....
Wes