Field sensors.

m_fabio2

Ok, here's a small project I want to complete, but I'm stuck

I want to be able to project the location of object within a given area.
Say the size a baseball field or a soccer field.

Couple of things I have thought so far:
A) Use some kind of radio transmitters at the four corners and meaure the time the signal takes to get to my object from these transmitters. Use this information to triangulate a location.

Any ideas ?

(My first thoughts were to do this so I coudl preprogram something say like a lawn mower, (I haven't addressed the multiple safety issues with this yet, but I know there are many)).

Post Edited By Moderator (Aristides Alvarez (Parallax)) : 3/16/2005 6:04:12 PM GMT

Robert Kubichek

How about multiple ultrasonic recievers and mesure the difference in time from a master clock signal?
You could use more than one at each corner, it depends how accurate you want it to be.

It woulld be cheaper-easier to do than a rf signal.

Bob N9LVU

Paul Baker

I don't think traditional radio triangulation would be possible with a stamp. Let me think about the idea abit to see if I can think of something.

I did my master's thesis working with high frequency audio communications (not ultasonic but it suffers from the same problems even more so). Emitters (speakers) become more unidirectional the higher the frequency, this is due to both the characteristics of the sound itself and the resonant material used to produce them, the object would require an·entire array covering the surface of the object. Our lab contemplated object location using ultrasonic but abandoned it because it was too directional and subject to being eclipsed by barriers. ULF (ultra low frequency) doesn't suffer from the same problem, except the diaphram required for it·is huge. one final problem with ultrasonic communication is that the higher the frequency the larger the absobtion rate of air is, for the distance he is talking about your likely to need an emitter of 200 dB or more.

RF is the way to go in general but the stamp is really slow. To give you an idea of what we are talking about, the RF will travel one meter in 3 and 1/3·nanoseconds corresponding to a sampling frequency of 300MHz

Post Edited (Paul Baker) : 3/16/2005 2:08:50 AM GMT

Robert Kubichek

Why don't you use a modulated light beacon, the detectors would scan a 90-180 degree arc depending on location in field and lock onto the modulated
light sorce. Then you could read the direction the detector is pointing to, and triangulate from multiple detectors. The more detectorsyou use,
the better positional accuracy you would have. This would be much easier on a stamp, as it would not require high speed.


Bob N9LVU

Paul Baker

Further perspective, RF will travel 25 km in the time it takes the fastest Basic stamp to execute a single instruction. With an SX RF will travel 6 meters for one SX intruction.

Paul Baker

Robert: you pose an interesting idea, it maybe too constrictive for his intended use.

Robert Kubichek

Sure, but I would use one stamp per detector, to lock onto the modulated beacon, using max detection level for direction, and send direction upon being polled to master stamp, this way each detector would have it's own id, and a person can use as many slaves as feasable or as needed. Also you can set the slaves to lock onto, and track the beacon since they know the modulated frequency. You could do realtime tracking depending on how often you poll slaves..

I would use a cat5 cable for power and coms to each slave, diasychained for simplicity.

You could do it with doppler rdf, but this would be MUCH more accurate, and lower in cost.

Bob N9LVU

Paul Baker

Ok the best I can think of is a transceiver on each end, where you relay the information back and forth enough times to be measurable by a stamp (I would suggest using an SX which would reduce the number of round trips required) you would have to keep complete track of the delay times used for the computational overhead for each trip and subtract that from the total value to obtain the travel time. The best would be to use 3 sets of tranceievers on different frequencies to prevent RF collisions between the different static units.

Paul Baker

actually 3 different frequencies isnt needed, the data packet would direct which static member its in communication with and it would proceed in round robin fashion with each one.

duckie

wow.. they sure do want ot drop some dough on this project.·

If I wanted to keep track of a lawn mower in a baseball field, I would probably use several base posts with dirrectional sensors mounted on a rotating base, and then a modulated light source on your target (lawn mower).· Think Flashing beakon on the lawn mower, and then toilet paper tubes mounted on servo's.· Look through the toilet paper tube and see the flashing beakon, then (based on the signal to the servo) yoiu know what dirrection the lawn mower is.· you could use a single sensor per base station (think old school radar, you get a blip when the sensor passes your airplane, but no distance information) or 2 sensors, per base station, and create a dead band between them.· WHen the right sensor sees the lawn mower, the servo turns right, until it doesn't see the lawn mower.· 2 stations make a triangle, a triangle tells you where the lawn mower is.·

Accuracy is going to be limited by how how small you can get your dead band, and sensing range will be limited by how bright your flashing light is.·

Of course, instead of a flashing light, you would probably want to use a modulated IR transmitter like a TV remote, and instead of toilet paper tube...· something easier to work with.·

You'll want a remoted lawn mower kill switch, and I fear for small children and rabbits.·

bishop

peh..rabbits.
i'd create a separate intake chute just for them.


muwahahaha....

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
**************

daniel woolston
Teksystems Inc.
www.danwoolston.com
**************

m_fabio2

hmm. If I can't run down childern and rabbits with this mower.. I don't see what fun it will be to get to work...(j/k)

Although I have a Engr. degree, I concentrated in high power and control operations, so as I understand the talk of using RF, implementing it is going to be a little difficult.

A friend of mine has a set of bincoluars that have a range finder on them. I belive it is some kind of infrared system. As I think about pricing and time invested, I'm thinking this might be the best route to go with this project.


Post Edited (m_fabio2) : 3/16/2005 11:27:59 AM GMT

Forrest

You could use a GPS and an RF transmitter attached to lawnmower to broadcast it's position, and an RF receiver to pickup the signal.

bishop

or you could use the buried wire technique as discussed in the robocut thread.
at least you'd have boundaries set. it wont help for guidance, but limitations will be set.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
**************

daniel woolston
Teksystems Inc.
www.danwoolston.com
**************

Forrest

This is an interesting topic - I really like Robert and Duckie's solutions. Assuming your have a base at each corner of the field, you could mount a pair of IR emitters/detectors mounted at 0 and 90 degrees at each corner, along with a visible light LED to make sure each base is lined up and square with the other bases. You could also mount an IR transmitter and receiver (at 0 and 90 degrees) on each base so each base could communicate with it's neighbors to triangulate the position of the lawnmower.

Paul Baker

Seems Fraunhofer and Cairos Technologies (both German companies) have teamed up to embed a wireless receiver in footballs (soccer to Yanks) to track thier position.
http://www.iis.fraunhofer.de/ec/navigation/tracking/index.html

This appears to be exactly what you want to be doing, but be prepared to whip out your German to English Dictionary, because produktbrosch

Ray Iddings

I say just spraypaint a black strip in a wwww form and have it follow the blackline attach a can of green to the back so it can paint the blackline over and voila your done. Cheap and Fun [noparse]:)[/noparse]

RickB

m_fabio2:

Check out this robotics site. www.convict.lu/Jeunes/RoboticsIntro.htm
Don't let the word convict throw you. Much of the navigation info on this site involves measuring time and phase differences and triangulation. VERY good stuff.

Rick