Private baby GPS

You need more than 4-5 transmitters for it to work decently, but TI's location engine fits the bill.

RF based devices will cover a larger area at a not so high accuracy. If the garden is not that large - or at the cost of more receiver stations - you can use a modulated ultrasonic transmitter to attach to the naughty kid and detect its position very precisely (at least until the kid does not get naughty enough to get rid of the tag... :-)

Dimiter

------------------------------------------------------ Dimiter Popoff Transgalactic Instruments

------------------------------------------------------

There are technological solutions to this.

Hmmm ... intersting application ;-) I'd like to be able to use it indoors. I am hoping to get half meter accuracy or better. The hints are a start ...

Any other takers ? :-)

What type of accuracy can be expected with ultrasonic ?

Regards G

I would expect that range measurements with an ultrasonic device would be on the order of a few centimeters.

However, using ultrasonic ranging over some tens of meters, you would have to cope with ranging errors due to wind velocity. Sound velocity in air is about 320m/second. Wind velocities could easily change the velocity by 2 to 5%.

Another problem would be the gain changes required to receive the pulse at distances from 1 to 30 meters. For an isotropic source, the energy received is going to vary ~ as the cube of the distance.

Mark Borgerson

A lot better than half a meter should be doable. At 300 m/S, 0.5m means 1.67 mS delay which you can measure using just about anything :-). Measuring 1/10 that may take more than "anything", but it will take me more than just an estimate to be able to say whether it will be significantly costlier than the "anything" :-).

Dimiter

------------------------------------------------------ Dimiter Popoff Transgalactic Instruments

------------------------------------------------------

Giorgis wrote:

I suspect limiting the range to something like 5 or 10 meters per receive station would be cheaper under the score than making them

Yep, that's one of the problems one will have to deal with.

Another one I thought of is Doppler effect related - what if the kid is very very naughty and runs lightning fast.... :-) :-) :-) (Sorry for not resisting that, I just pictured it and got carried away :-).

Dimiter

------------------------------------------------------ Dimiter Popoff Transgalactic Instruments

------------------------------------------------------

Mark Borgers> > >

Arn't the ultrasonic sensors directional ? If you want to find how far away a kid is, don't you kneed to know were it is What about it's body obsuring the line of sight ? What about signals boucing around ? Wind speed was mentioned, but what about temperature effects ?

Laser is alot more accurate I guess, but directional.

It's would be great if we could hi-jack GPS chips for down to earth sensing.

Maybe we shouldn't use kids, it's getting creepy. What about fluffy kittens ?

It's all about speed and timing accuracy. Since sound travels about a million times slower then radio waves, you can expect a much higher precision with (ultrasone) audio then with radio.

A basic (and maybe naive) setup could be to combine RF and audio: your device-to-track transmits an audio and RF burst at the same time, a few times per second. These signals are picked up by 3 microphones and a radio receiver in the field. Calculating the time between the incoming RF and audio signals should give you enough information for making a 2D fix of the position.

Hmmm ... nice ! Because ultimately I don't want a returning signal to measure distance.

Can I ask another, is it possible to detect the angular direction of the source fo an RF signal with a MEMs device ?

MEMs ?

Flux gate compass maybe. It's not a MEMS device.

I see. I don't think something mechanical would be appropriate here. Maybe something can be deduced from phase difference on multiple receiver antennas ?

Hi Giorgis,

I think what you are looking for is a LPS (Local Positioning System) Take a look at

Sandeep

That would be rather wasteful. It's certainly easier to install a 4th audio receiver to resolve time on top of distance than to bother with a RF transmitter/receiver system on top of the ultrasonic gear. If this is going to be a wannabe GPS, it should mimic the original as close as possible, i.e. use only one signal channel.

A more useful variation of the original principle might be to invert it: have the tracked device send 'pings' at arbitrary times, and collect arrival times from 4 sensors spanning a 3D space. Then break out the maths textbook to find out how to turn that into a positional fix.

Hello,

perhaps a completely different idea: how about a camera high above, which extracts motion / changes against background ?

Regards,

Martin

The setting is just an example. I simply would like to duplicate the consept of a GPS but on land and over radius say ... 30m

Giorgis

If you are really intent on duplicating the GPS concept, where you have several transmitters at known locations and a receiver at an unknown position you will have the following characteristics:

1. The distance from transmitters to receivers varies by only a factor of 2 to 3 (a guess based on satellite orbits and ground distances). That means that you will have to locate your transmitters a few hundred meters outside the sensed area.
2. Timing of the pulses from the transmitters will need to be pretty well regulated (1 nanosecond timing error ~= 1 foot distance error). That means very good clocks or hard wiring to the transmitters from a single source.
3. The receiving unit can figure out it's position, but that is of no help unless you can either log the position or transmit it to the interested party.

Mark Borgerson