Finding a rabbit with a PICAXE...

J

Jeremy Leach

Senior Member
Here's a little discussion of using radio etc <A href='http://www.ece.stevens-tech.edu/sd/archive/01F-02S/websites/grp15/public_html/ProjectProposal.html' Target=_Blank>External Web Link</a>

And a very good summary of techniques here <A href='http://www.doc.ic.ac.uk/~nd/surprise_97/journal/vol4/jmd/' Target=_Blank>External Web Link</a>

DPAG ...I think you said you didn't get round to trying it. Unless you've done something very special I can't see it easily being a workable/accurate solution.

Edited by - jeremy leach on 08/02/2007 20:59:51
 
B

BeanieBots

Moderator
Thanks Jeremy. The articles highlight my point.
a) Sub nano-second timing requirements.
b) Poor resolution.
c) Dosn't work in confined areas.

Not really something for the home constuctor (RF triangulation). Even if the budget does include your own satalite, the resolution is still not good enough to locate Nesbit in her garden.


Edited by - beaniebots on 08/02/2007 22:17:40
 
F

Fowkc

Senior Member
This is why I think ultrasonics would work better.
Time for a radio wave to travel a meter: 3.3ns
Time for a sound wave to travel a meter: 3ms

I think with a bit (OK, a lot) of work, ultrasonic triangulation is more than possible.
 
I

inglewoodpete

Senior Member
Take a leaf out of the book the people that study animal movements use:

1. use a radio collar and a dipole antenna receiver (highly directional)

2. use GPS technology on via satellite link.
(works well on whales)

Actually, option 1 may work with two receivers physicaly scanning (panning) the target area.

My preferred option (failsafe):
- Araldite a GPS to a heavy rock.
- Tether the rabbit with a metal cable or chain (lets say length = 1 metre) to the rock (rabbits chew through anything else, making readings unreliable).
-Get a GPS reading of the rock.
-Rabbit will be within a metre of the coordinates of the rock.

Got to be honest: I haven't studied rabbits much, except through the sights of a rifle.
 
D

demonicpicaxeguy

Senior Member
unforunatly it never got past concept
because the cat died
but the idea is to have the output of the reciever going directly into the transmitter on the collar so it just repeats what it recieves

then at three different points in a triangle
each point takes it in turn transmitting a pulse which is recieved by the collar and re-transmitted, the point recieves the pulse a very very short time later then measures the delay from that you'd be able to work out how far away from each point the cat is

at the time i was simply going to use 3 tape recorders and a 555 timer and look at the data later on a pc
but these days if a picaxe's pulsin can measure a small enough pulse then it should work... i've been thinking about it for a while and have been contemplating doing it with my kid(S) while we go shopping as one of them has a tendancy to wonder a little

personally i don't think it will work
due to accuracy small timing values etc
it may over a very larger distance

i have no doubt however it would work very well with ultrasonic transmitters


Edited by - demonicpicaxeguy on 09/02/2007 05:43:19
 
M

moxhamj

New Member
So many helpful suggestions??! I was going to suggest planting a carrot patch and looking in there for Nesbit but then I noticed that Nesbit prefers cabbages. I have lots of wild rabbits round my house and a completely untouched carrot patch! Then last week I met a brown snake in the carrot patch and it all became clear. At least I know my vegies are safe from rabbits.

Yesterday a patient of mine came in and told me about his new job soldering up circuit boards for animal tracking devices. Link is http://www.btms.com.au/ click on Tracking then Product Specifications. He says the smallest animal they track is a hopping mouse, and that is with a full GPS/directional system. I'm sure you can find a local supplier of similar systems.

Edited by - Dr_Acula on 09/02/2007 05:45:02
 
D

demonicpicaxeguy

Senior Member
or a can of glow in the dark green spray paint...
the only place to hide then is underground

Edited by - demonicpicaxeguy on 09/02/2007 07:51:50
 
D

Dippy

Moderator
Demon. You'd never have got your time-delay thing working with standard stuff. The time delays would just be far too short for you to measure. How far does light travel in a microsecond?

It should be noted however that SOL time delays can affect maximum ranges of some GSM systems (base to phone).

U/S would probably work, thats I (jokingly) suggested it earlier. But the practical side in this application would render it useless.

Anyway, this has been an entertaining thread. Time gentlemen Please. I'm rolling the pastry now.
 
D

demonicpicaxeguy

Senior Member
acording to wikkipedia the speed of light is 299 792 458m/s
and the smallest a picaxe can measure is 2.5us so...
that would mean our signal would take 1 microsecond to travel 299.79 metres per microsecond
now at 2.5us resolution that makes a distance 749.475m
that makes it a too big an area
so we need to be more realistic nanoseconds!!!!!

it take 3.3ish nanoseconds for light to travel 1m

so to get a resolution reasonable to find a rabbit 33.33333 cm is achievable at 1 nanosecond

measuring that maybe somewhat problematic.. me thinks.. it seems timers that come even close to that are more military use kind of things

i don't think texas instrumnets would send that sort of thing out as a sample

it seems a lot of trouble to find a rabbit in a backyard

however i think it is still possible to some extent and i'm going to have a think about it



Edited by - demonicpicaxeguy on 09/02/2007 10:34:28
 
D

Dippy

Moderator
And have you looked at the response time of RF receiver modules? (off the shelf receivers)
Forget it.
 
D

demonicpicaxeguy

Senior Member
since we can't get a timer ic that does down to 1 nano second how about this?

simply bounce the pulse between collar and point A while doing this

(assuming you can get one that does up to 3 ghz) use a frequency counter to measure the freqnency of the pulses

obviously the further apart the collar and point A are apart the lower the frequency will be.
there will be some delay within components interference etc

i probably could come up with a 100 different reasons why it wouldn't work
but i think the theory is sound... to a point... it's got me interested now
i'm fairly sure there are 3ghz frequency counters availble i'll see what there is and do some more maths


 
B

BeanieBots

Moderator
&quot;obviously the further apart the collar and point A are apart the lower the frequency will be&quot;

Don't see how the frequency will change?
(unless Nesbit is hopping at very high speed!)
 
F

Fowkc

Senior Member
Dippy: what do you think the problems are with an ultrasonic system? I've never used ultrasonic stuff before (well, not a lot).

I can see directionality being a problem, but perhaps this could be mitigated enough by putting the transmitting mics as high as possible in the garden, and point the mic on Nesbit vertically upwards?
 
J

Jeremy Leach

Senior Member
Ultrasonic ...

* This is on a Bunny not a robot, so very unpredictable orientation.

* Reflections of sound

* Obstruction of 'line of sight'

* Power consumption

* Bunny can hear ultrasound (?)

* Cost

To name a few problems ....at a guess.

 
D

Dippy

Moderator
Oh dear, this is all getting a bit silly.

You won't be able to buy a radio receiver (well a transponder really) that could operate fast enough. Maybe the military boys or aircraft people can do it, but they are talking much greater ranges and longer delays. No doubt someone has a cousin in the army and will tell us all about it. Yawn..

Yeah, you can get fast counters but its the receiving bit where it will all fall down. The response times of your average cheapie receiver are thousands of times too slow for what you want. Don't waste your money.

Frequency changes? BB right on. The only way you can really detect this is via Doppler radar (which coincidentally I'm working on at the moment using FFT). Over short distances actually measuring distance is a no-hoper. I can't afford Ground Radar.

Ultrasonics? Just as Jezzer has said.
U/S transducers have a certain angle of dangle, so you'd need to rotate or have an array. The propogation is slow enough for practical experiments, but Jeremy has highlighted the problems.

On a rabbit, forget it. I only made the suggestions because I assumed everyone else was joking and having a bit of a laugh. Its certainly brought a grin to my rosey cheeks.
 
hippy

hippy

Ex-Staff (retired)
Indeed &quot;silly&quot; in places, but I think it's been a worthwhile and entertaining thread. In terms of 'brain storming' and exploring a whole range of [im]possibilities.

Dippy : Took me a while to understand your comment about 'rolling the pastry'. You naughty, naughty person :)
 
T

toxicmouse

Senior Member
presumably you fetch Nesbit after work, at dusk or night and allow your pet to roam your garden during the day?

what about a 08M, an LDR and an LED. LED flashes when there is low light intensity. spice this up with a receiver instead of an LDR, then you can impress your friends with a glow-in-the-dark pet (the ultimate in gadgets). for those of you down South- makes an ideal Tokolosh!
 
R

Rickharris

Senior Member
OK lateral thinking here - how about a Tx that sets an LED beacon when it gets near. After a few seconds/mins the LED resets - that way you can see where he has been and following the resetting LEDs get a line on his direction - Assuming not too much movement.

this way the bunny carries a small tx and all the hard work is in fixed beacons. this could work with Ultra sonic or IR as well rather than radio.

Edited by - rickharris on 09/02/2007 19:55:02
 
D

demonicpicaxeguy

Senior Member
what i mean is if you were to have a pulse bouncing back and forth continously
as the rabbit moves around the garden

(eg the output in each reciever going to the input on the transmitter and given a short pulse)

the distance between the three points would change therefore the time it takes the signal to bounce between collar and points so the frequency would drop the further away the rabbit got
i'm simply saying if you could accuratly measure the frequency you could measure how far away the rabbit was

i'm not talking about the frequency of the transmitters i'm talking frequency of the pulses getting transmitted
 
B

BeanieBots

Moderator
You mean:-
a) send a pulse (+/- time variation of sending trigger, plus Tx start time)
b) receive the pulse (+/- time to detect pulse, plus Rx response time)
c) send pulse back (error as in a)
d) receive returned pulse (error as in b)
go back to a)

and meanwhile measure the period between pulses (frequency) to the nearest nano second.

I don't think so!

Rick's idea has some scope. Was thinking along the same lines but using low frequency RF and some sort of ground loop. Similar to the method used by industrial robots to navigate around factories but the other way around.

Edited by - beaniebots on 09/02/2007 20:57:45
 
D

demonicpicaxeguy

Senior Member
ok given we can't time down to the nanosecond
approach two.....

there is somthing we use at work called a correllator (i look/listen for water leaks in sydney water mains)

eg
www.primayer.co.uk/leak_location_eureka.htm

what they simply do is digitally vary the delay on the two channels and draw a percentage graph on the screen on how much the sound matches between two channels

then using some basic maths after being told the composition of the pipe and size along with total length it can work out a distance from red or blue...

for what they do and how they work they are grossly over priced (we get them for 20grand Aus) but the concept on a smaller scale with a picaxe shouldn't be too hard
 
R

Rickharris

Senior Member
May as well go for brain waves Fowkc, Ah Rabbit - No brain!! <img src="smile.gif" width=15 height=15 align=middle>
 
W

Wrenow

Senior Member
Fascinating and fun thread.

I am kind of surprised no-one appears to have suggested a phased array in two corners of the yard, each generating the angle from which it receives the signal, and triangulating the bunny's TX from the two angles. Or perhaps a couple of base line interferometric arrays. See http://www.drdo.com/pub/techfocus/apr2000/Direction%20finding.htm
Of course, you are then getting into some esoteric RF stuff (which is well above my paygrade), but it is a road well traveled by the RF direction finder guys, and has been around since WWII (check out the HF/DF aka Huff Duff on Google, and see how it helped defeat the wolf packs).

Of course an R/C beeper and strobe would probably be the easiest and cheapest - no picaxe necessary. Just click the &quot;Find the Bunny&quot; button, and look for the light/listen for the beep. Unless the garden is really large.

Just my .02, YMMV.

Wreno

Edited by - Wrenow on 10/02/2007 00:05:15
 
hippy

hippy

Ex-Staff (retired)
Then there's thermal imaging CCTV, maybe even normal CCTV ( IR for night ) with some motion detect software. Nesbit might be in the undergrowth, but tracked motion when not should allow the trajectory to point to the likely location.
 
J

Jeremy Leach

Senior Member
Any bright-ideas here are really much broader use than detecting a Bunny.

I think Rick's suggestion of detecting proximity to beacons has a glimmer of possibility (nodes might be a better word here, because I see the unit on the Bunny as the beacon).

However I can still see many issues (signal strength varying enormously making nodes 'trigger' erratically, plus if trying to use IR, swamping of signal by sunlight (S/N Ratio) ...etc etc). Plus the cost of the intelligent nodes.

The link mentioned by Dr_Acula shows how the professionals do it using directional RF (plus using automated direction finders), but the .pdf on the website shows that it's not exact.

The links I mentioned earlier really did spell out a lot of the issues with the different techniques.... but who's to say there isn't another bright idea no-one's thought of!

Just for a laugh ...I thought of a grid of very low wires with sensitive microswitches at the ends to detect when bunny brushes past and so can locate the 'cell' shes in....except these 'trip wires' would knacker you coming back late at night ;-)



Edited by - jeremy leach on 10/02/2007 08:34:12
 
F

flyingnunrt

Senior Member
Hey guys jsimpson hasn't been with us since page 1..., but for the rest of us go have a look at what Roboteernat posted on 16/1/07
 
B

BrendanP

Senior Member
Put a small RF reciever on bugs (radiometrix makes a unit that is for fitting to migratory birds). Transmitt signal to Rf unit which sers'in by 08 and then sounds small buzzer on bunny.

Then for a week or so always feed rabbit in his hutch and before and while he is eating a nice tasty sort of food trigger the buzzer regularly.

Very soon the rabbit will make the association of the buzzer with the food in his hutch and come running.

Use some sweet food like molasses sweetened horse food.

I dont know if this is what the guy wanted but I think it will work.

I enjoyed reading the other posts, there are some smart guys here.
 
J

jsimpson001

Member
Hey guys I am here, some really good suggestions (if a bit on the daft side sometimes) :) what brendanp said sounds like a good idea.

A directional sensor placed in the middle of the garden would be ideal; when i wanted to find Nesbit, I could 'activate' an auto sensing antenna, which begins to rotate and stops when the 'loudest' signal is picked up, then lights an LED in this direction (presumably with a bipolar ant this would be 2 points 180deg apart) two identcal sensors in different places would solve the problem which you just thought of when i mentioned the two points.

or I could just take the really simple approach and go for the 'rabbit trained' idea that brendanp suggested with 2 units, one on the Nesbit's back, which waits for a signal (RF) from the handheld unit, when it recieves a signal ,it produces a loud buzz and ultra bright LED flashes for a period.


moving over to ergonomics for the rabbit unit, im thinking of an elasticated strap around the belly, but this doesn't stop the unit sliding underneath the rabbit.


 
R

Rickharris

Senior Member
<BLOCKQUOTE><font size=1 face=arial>quote:<hr height=1 noshade>Nesbit is ok with that, she sometimes wears a lead. <hr height=1 noshade></BLOCKQUOTE></font><font face='Verdana, Arial, Helvetica' size=2>

If you make the LEAD a big lump she won't go anywhere! <img src="smile.gif" width=15 height=15 align=middle>
 
J

joe of loath

Member
what about a rotating dish and a rabbit pinger? the dish picks up the ping and tracks it, giving you a rough estimate of where the rabbit is in the garden.
 
You must log in or register to reply here.
Top