SparkFun Forums 

Where electronics enthusiasts find answers.

Have a good idea for a new product for SFE or Olimex? Let us know!
By Highway
I'm an IT professional with some land and some livestock. I am a part time farmer, but am very part time, and am looking for ways of saving time in every aspect of my farming. One of the most time consuming parts of the job is counting the livestock, and just generally monitoring them. Sometimes you just won't notice problems unless you just watch for a while. Good animal husbandry is not something you can rush. However, it is something you can do more efficiently I think!!

Being a bit of a techie, I'm interested in ways of improving the efficiency of this process, to free up more time to do a bit more "productive" farming.

I know absolutely nothing about electronics, but here is my idea:

Produce a small weatherproof device that is strapped to the animal and does the following:

1. power/charge itself from the movement of the animal, like a kinetic watch that needs no batteries (or at least only needs batteries change every few weeks)

2. send the coordinates of it's position at preset intervals. I'd like a google type map of my farm with dots representing each animal. I'd like to be able to calculate how far/fast each animal has moved, with warnings being flagged if an animal has not moved for a long period of time, or is constantly moving rapidly (both bad signs), and needs attention.

I'm thinking/hoping some sort of gps coordinates could be transmitted in some way ultimately to a server which will plot them on a map, and process the data to generate any warnings, etc.

3. Minimum running costs (e.g. network charges , etc)

4. Minimum maintenance. Little use if i have to catch every animal every few days to change batteries, manually download info, or whatever.

I have seen various products on your site that might be combined to achieve this (e.g parcel tracker is a good start).

So i guess I'm wondering if my wishlist is in anyway achievable or realistic, and if anyone has suggestions on how I might go about it.

Even suggestions of books on electronics to help me understand my options would be useful.

There are commercial options available, but they are expensive, and thats no fun!

Many thanks,
By theatrus
Highway wrote:Hi,
1. power/charge itself from the movement of the animal, like a kinetic watch that needs no batteries (or at least only needs batteries change every few weeks)
There simply isn't going to be enough power generated to be useful here. GPS units and any long range wireless devices are also power hogs.
Highway wrote: 2. send the coordinates of it's position at preset intervals. I'd like a google type map of my farm with dots representing each animal. I'd like to be able to calculate how far/fast each animal has moved, with warnings being flagged if an animal has not moved for a long period of time, or is constantly moving rapidly (both bad signs), and needs attention.
This is the tough one.

Your options are:
1. Cellular data. High power. Needs service on your land, plus large cost PER device. Bonus point: if someone steals your sheep, you may be able to track it if the collar is still attached ;)

2. Wifi. High power. Really depends on your geography. Complex software requirement.

3. Low power mesh network. Complex to design and implement. May not be a complete mesh (not all animals reachable, not reachable to the base)

4. Higher power "mesh" (the over 1mW transmitters). A compromise between Wifi and low power mesh networks. Depends on the geography and distance. Probably the cheapest route. Use them in a simple star configuration for simplicity.

Battery life with GPS and Wifi is always a concern. The upside is you may be able to make do with minimal reporting times (take a snapshot once an hour for instance) which will help a lot in this area. Depending on the chosen technology, and hardware design, you should be able to get month+ from many power sources.

I work a lot with very low power devices at the "day job", and get year+ power from AA cells using low power mesh networks (no GPS of course). Your enemy is distance and wireless.

You'll probably need firmware authoring expierience for all of this.
Last edited by theatrus on Thu Oct 02, 2008 4:15 pm, edited 1 time in total.
By Highway
Thanks very much for the very prompt reply.
I figured that power would be the issue, but you've given me something to go on at least.

By Philba
The power issue might be surmountable. Duty cycle is your friend here. Basically, if you only turn the system on once an hour for 2 minutes then you really need to generate 1/30 of the draw. In practice, you will need a somewhat higher current due to losses but we can ignore that for the first cut. So take the current draw of the devices, figure out how long each device needs to run and then do the math. Once you've computed the total current draw then you can determine the average need.

Next you'll need to determine the how the generator is going to work and how much it's going to generate. I can't help you on that - maybe you need to experiment. A quick google of linear generator has some interesting hits - mostly for wave energy generators but the concepts are the same.

By the way, the GPS units look like they won't pull that much current (~30 mA at 3.3V) and it looks like they will run for less than minute to get a lock. Not cheap though.
By Highway
Thanks for that info. I think I'm out of my depth, but I'll look into the generator and see if i can figure it out.

By UofMEShop
Recharge with solar ?
By Highway
I was thinking of something like that myself, but don't know enough about it.

Here is a crazy idea! Is it possible to setup my own access point, as a mobile carrier would use, and buy and program blank sim cards to communicate with that? Probably need licensed frequency, and a lot load of stuff.

Nevertheless, if i had a box that has a gps unit to gather location/speed, etc and solar panel to keep the battery charged, and a private apn that communicates with diy sim cards to upload the data to my webserver, all in a weatherproof box.

By lyndon
Interesting idea.

Is there any reason you're not using a sheepdog? I also live on a farm ( -- we have chickens and horses) and I have the same thoughts. It's like a second job and anything you can do to save time is worthwhile. However, sometimes the old ways work best :-)

My thought would be to have small RFID units on each head of livestock and if you know that your dog (if you have one) will herd them past a reader once a day, that will take care of the counting problem. As far as animal health goes, it might be simpler to have a few remote solar powered WiFi cameras that can be panned to look for downed sheep or predators.

It is an interesting idea: animals are always escaping their confinement and this would be one way to know before they got too far away and hurt themselves. I'm sure other farmers would be interested.
By tah997
as lyndon suggested, RFID may be your best option. I have seen similar systems used in Australia. All of the station's cattle are individually tagged and positioned at certain gates around the property there are RFID tag readers. whenever a head of cattle passes the reader, its position is recorded.

The property has a large database containing the age, date of purchase, weight, vacinations and bloodline of all their cattle. I guess it helps them sort out which cow/steer goes to the abattoir today and which one goes next week ;)
By teekay_tk
How about using GPS animal trackers, they are expensive to start with but running cost might be low and they usually run for months some time for years.

we make thse for marine mammals ... px?psr=287

other companies that make such technology are bluesky telemetry in UK, wildlifecomputers in USA and Sirtack in NewZealand.
By Highway
Thanks for all your answers. I'm encouraged that my idea is not considered daft.

I don't have my own dog at the moment, (use my dads when needed, but he usually ignores me, the dog that is!). I had thought of running all the stock through a gap with a reader of some sort positioned there, but having to go to the field with the dog and round up the livestock (the walls are currently bad, and they can roam considerably) to run them through a particular gap/gaps would probably not take much less time than just counting them.

When I do get the walls repaired and am able put in a proper rotational grazing plan then I think this simple idea would work fine. However, if i did manage to get something working for myself this way, it definitely would not work for my Dad's land due to it's layout, so I might as well try and do something we can both use.

The wifi camera idea is a good one too, I hadn't thought of that at all. Again though, with the land layout, I'd need a lot of them, and my Dad would need a trailer load.

What might work with the RFID tags idea is to simply drive around with a
reader on the tractor, and drive within comms range of the flock. I could then possibly analyse the data right there on a laptop (or a smartphone or whatever) and get an immedidate count and any suspicious stats. If there is one missing, the system would know which one and where it last was seen. This could be the most cost effective.

I would not mind the driving part so much, just not having to do 10 laps to get the count right and then sit there watching any unusual behaviour for half an hour.

Teekay_tk, your solution is precisely what i am looking for, but is way outside my budget unfortunately. Now if it was possible to make a diy version, with a private mobile apn or something, etc, then I'd be sorted!
Would you be interested in such a project?!

By foxxjnm
You could definitely something similar to one of the post above where you have readers in various locations on you land (close to a water source is a good place) and use a GM862-GPS in combination with an RFID reader. Check this link out: good luck.
By henryhallam
Here's a more out-there idea:
Put a camera on a tethered balloon. Let it rise up to a few hundred feet, take a photo of the field and have a computer vision system count sheep.
By RealLife
henryhallam wrote:...have a computer vision system count sheep.
Wouldn't that put the computer vision system to sleep? ;)

Sorry - couldn't resist.
By 48X24X48X

I guess you can utilize the mesh network concept to help localize the sheep locations in the farm. You can place higher power radios on stagnant position. Whenever the sheep moves a particular fixed higher power radio, it will be identified to be around this fixed radio location provided that the sheep is carrying lower power radio version. The receiver at your home can then do a route check on where the data came from and from which sheep. Using low power module (0 dBm), a coin cell may last you a year or so. But the fixed radio probably needs a small size solar. This method might not be super accurate in terms of positioning but provides location approximation up to a certain degree. If you want it to last slightly longer, I saw some very very tiny solar cell (as small as 0603 SMD resistor), can series a few of them in order to have enough current to charge the batteries and Sparkfun carries the rechargeable coin cell too!

We tried to do horse tracking last time for a horse race event using similar approach. The only different is sheep moves slower than horse. :D
long long title how many chars? lets see 123 ok more? yes 60

We have created lots of YouTube videos just so you can achieve [...]

Another post test yes yes yes or no, maybe ni? :-/

The best flat phpBB theme around. Period. Fine craftmanship and [...]

Do you need a super MOD? Well here it is. chew on this

All you need is right here. Content tag, SEO, listing, Pizza and spaghetti [...]

Lasagna on me this time ok? I got plenty of cash

this should be fantastic. but what about links,images, bbcodes etc etc? [...]