[aprssig] rfid uses in real world.

Wes Johnston, AI4PX wes at ai4px.com
Tue Feb 23 13:40:28 CST 2010

Every year there is a huge motocross race here in the forest.  400 riders.
I had thought several years ago about using rfid for them.

What would be cool would be if the RFIDgate had two modes.... static table
and moving object.  Bob showed us what a static map of rfid woud look like
last week over the top of the dayton arena.

What about moving objects?  I remember years ago Bob mentioned about how to
track runners without a GPS.  Bob said to create an object moving at 6mph
for the lead runner and 4mph for the slowest runner.  Why can't the RFIDgate
do just that?  We know the average speed of the motorcycles should be in the
range of 25mph, so why not have a provision in the gateway to publish
ojbects with a bearing to the next checkpoint at xx mph?  This would still
give us the spacial diversity we're looking for in the static table made by
diddling the hundredths of a minute.  In this case we dead recon an object
instead of mucking with the lat/long.

Now let me take this one step further.

In the case of our motocross race, the course is 5x13 miles of forest, but
it's over 70 miles of trails.  These trails crisscross each other like a
tangled string.  To see a rider's location on a map doesn't really tell you
how far he is thru the race (unless you have a GPS breadcrumb map overlaid)
In the end, people just want to know how far the riders are on the course...
in linear terms.  Rider 1 is 50% between checkpoint 1 and 2 for example.  So
why not create a pseudo map that contains checkpoints vertically down the
left hand side with horizontal lines leading away to the east.  So as a
rider crosses check point one, we move him to the position of the checkpoint
(on the map) and give him a velocity of 25mph @ 90°.  In the 15 minutes it
takes him to get to the next checkpoint, his icon will have dead reckoned
6-1/4 miles due east.  When he gets to check point 2, we reset his lat/long
to the QTH of checkpoint 2 and give him a speed and bearing of 25mph @
90°<25mph at 90°>.
We have now  provided the audience with a linear map of the course.  They
can now see how far along completing the course a rider is.  We'll end up
with x number of rows of motorcycle icons dead reconing on the map.
This line segment table needs to be "located" over a vacant area like a
nearby lake.

Another perk of this is that riders who don't make it to the next checkpoint
just keep right on deadreconing to the east right off the end of the
course.  So now we know at a glance that we have some riders who probably
broke down between checkpoint X and checkpoint Y.  We no longer have to wait
for all the riders to get back to the start to begin looking for the
breakdowns and medical cases.

As far as the RF logistics go, we'll have too many riders and packets to fit
in the 1200 baud channel.  Figure on 5 checkpoints in a 2 hour race, 400
riders and you've got to transmit 30bytes 2000 times in 4 hours.  We'd have
to run each check point on a different frequency back to the start of the
race (or commo trailer).  Then we can MUX all this data into one stream
using a copy of xastir running a telnet port.

What you think?
God help those who do not help themselves.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.tapr.org/pipermail/aprssig/attachments/20100223/e7a25095/attachment.html>

More information about the aprssig mailing list