Go Down

Topic: Outdoor tracking (Read 5937 times) previous topic - next topic


I would love to see a practical, real-world example.

Well, WAAS has already been mentioned.

Differential GPS does the same thing, on a smaller scale.  It uses a fixed station at a known point which calculates errors and send them to the mobile unit for corrections.

An obvious example would be one of those spiffy military GPS units, like those used on JDAMs, cruise missiles, etc, that use the second (still encrypted) GPS frequency that's far enough away from the ~1.5Gz signal we mere mortals use that the atmospheric distortion can be compensated for onboard.




Get two of these and xbee's, you can easily have your cart follow you.


So you think that theese gps units are precise enough relatively??

What is the purpose of the xbees?


Ok i wasn't totally awake when i wrote this.
I was still focused on "ir tracking" but using gps the cart off course need to know where i am ;)
This is an expensive solution, i need to know if it is stable enough.


Sep 23, 2010, 01:46 pm Last Edit: Sep 23, 2010, 01:46 pm by kg4wsv Reason: 1
Perfect. All Mr. Pulse has to do is increase his GPS budget by 20x

I'm not sure of the cost of a DGPS rig.  I'm sure comercial is expensive because of its applications, but I don't know if one can be homebrewed.

and then have his robot and the target both stand stock-still while they make a precision fix.




A cheap and quick solution:

A thin rope attached to you, and to a joystick on your bag.
When you move it will pull the joystick on your direction and then let know to the Arduino where to go.

Another way could be a 360° coverage based on ultrasonic long range sensors + a button to start and stop the stracking (and avoid your bag to follow a rabbit crossing arround)


Sep 24, 2010, 08:45 am Last Edit: Sep 24, 2010, 08:54 am by ThePulse Reason: 1
Hi all

I was on an electronics exhibition yesterday and got an explanation on how it works!
Before the explanations I was not totally aware of the difference between absolute precision and relative precision!
Well! Error correction, by having a fixed reference point, works because the same error is shown in both devisees and therefore you can correct the data on the moving device. Because the same error shows up un both devices, then there is no need for error correction, because gps1 has to follow gps2.
So the absolute precision is "gps to satellite"
and relative precision is  "gps to gps"

So if i bought 2 gps units that can have a DGPS corrected precision of 1 cm, then the "golf bag" would follow me around with a precision of 1 cm.

But it is still expencive!
For this solution i need 2 arduinos, xbee tx/rx and 2 GPS units!

And for an ir solution (if it is possible) i only need 1 arduino and some ir tx/rx


1cm precision can't be reached for civil GPS devices. 10m will be the best you can have.
So your bag should be somewhere 20m away from you, and it will be funny to see you trying to catch it when it's thinking you'r always 10 or 20 m away... ;D


@razorbob:I've seen civilian-grade GPS antennas attached to opposite ends of bulldozer blades for precision grading. Do you think that would work if they were only accurate to 10 metres?
Per Arduino ad Astra



You are right about absolute positioning to the world coordinates!

But the thing in this case is that 1 GPS unit is telling another GPS unit where to go!

So in theory if there is som kind of atmospherical distubance that makes GPS1 jump 5 meters in a bearing of 200 degrees, then the same error would show up un GPS2, so that the relative position between the 2 GPSs are the same!

Differential GPS (DGPS) works by having a GPS unit at a known position (lets say 100,100) but this GPS reports  101, 98) -> then this error (1,2) is sendt to the moving GPS unit out in the field by gprs (or other way) and applied to the GPS unit as correction data!

This can ONLY work because the 2 GPS unit sees the same (absolute) error!

Remember, i don´t have to know where i am in the world, i "only" a relativ position between 2 gps units that are acting on the same data!
So the important data to look for in the GPS unit is not the "absolute" value, but the DGPS precition -> because this value is showing how precise the GPS is itself!


Sep 24, 2010, 12:57 pm Last Edit: Sep 24, 2010, 12:59 pm by ThePulse Reason: 1
IR test!

I just performed an ir test in sunlight using the arduino, a standard remote, a buzzer and a TSOP85 IR Receiver!

As you would imagine the sensor gets saturated from direct sunlight (not reacting to the remote at all)
Holding my hand in front of the sun( to provide shadow for the sensor) from a distance of 4 meters worked!
So building a small "roof" over the sensor will do the trick! (of course it cant drive directly in to the setting sun but.......

It could be interesting to try to put an UV filter in front of the sensor to see if that would do the trick with direct sunlight, but i havn´t got one at the moment!

The IR transmitter wasn´t affected by the sunlight at all!


It could be interesting to try to put an UV filter in front of the sensor

Interesting, but largely pointless - the IR sensor isn't going to be very sensitive to UV (opposite end of the visible spectrum), so removing UV isn't going to make much difference.
Better to use a notch IR filter, matched to the peak sensitivity of the photodiode.
Google "Wratten 87 series".
Per Arduino ad Astra

Go Up