6

Is there a way if I have a paddock 2 km by 2 km I can find the location of my livestock, essentially live tracking their movement?

I've thought of possibly giving them a passive RFID tag with antennas around the perimeter but that would only work if they cross the fenceline boundary which would be more like checking what paddock (zone) they're in rather than where their exact location within the paddock? Also the antennas aren't capable of sending a signal long enough across a 2km perimeter right?

Is it possible to triangulate their position with multiple antennas pin-pointing their location?

I've tried to do some research but struggling to find a solid solution that isn't ridiculously expensive.

6
  • 2
    I found this article that could be interesting to you : iosrjournals.org/iosr-jce/papers/Conf.17025-2017/Volume-1/…
    – YCN-
    May 22 '18 at 15:19
  • Just a terminology note from a former rancher... something that size is not a "paddock". It's a "pasture". May 24 '18 at 23:08
  • Could you tell more about the reason for not wanting GPS? Too expensive? Too power-hungry? No sky view from the mounting point?
    – Sylvain
    May 28 '18 at 18:45
  • too expensive, ideally wanting something slim and disposable like an RFID tag on plastic admission wristband and certainly too power hungry for my liking. I just thought as I have a 'zone' rather than anywhere there would be a way to put antenna's on the borderline and triangulate within the zone. @Sylvain
    – Angus Ryan
    May 30 '18 at 4:22
  • When you say "tracking their movement", what does that mean exactly? Do you need to know where they are at certain specific times of day (e.g. at 12 noon and 4 PM), or do you need to know where they move minute-by-minute during the day? Do you need to know the movements of and track each specific animal by some sort of ID? What is the accuracy required?
    – Seamus
    Jun 6 '18 at 1:11
6

I think the least expensive technology that exists in this arena is active RFID. But the definition of expensive is best left to the user. A company called Monnit has active rfid tags and sensors directly targeted at livestock assets. Still only has a 250 ft NLOS range, but it's hard to know what the LOS range is without putting a sensor up a pole. Since the furthest location from the fence is 3300 feet. . . this probably isn't the answer. But I do think it's the state of existing technology.

In a somewhat longer, hackier timeframe, this looks like a fine application for a private implementation of Narrowband IoT in an unlicensed frequency.

First person to make this work could make a fortune. I don't know how many cows you have but I suspect there are those with many more who have the same problem.

2
  • 1
    Indeed, if you find a way with a passive device to make a good tracking system (at low cost), with only one active antenna, that would be amazing. But with the current status of tech I don't see how you can do that at low cost. You probably can find some hacks, but I feel like the only hardware low cost would be to use a node of camera with an openCV server and tags on the cows. But you still have the camera node battery problem...
    – YCN-
    May 22 '18 at 15:10
  • 1
    It's funny you went there because I was wondering whether computer vision would be the elegant solution. The big issue there seems to be identifying each individual cow. Holsteins generate their own ID code but monotone cows won't be so easy to identify. Common cameras and a PoE should solve the data capture side cost effectively enough, and it's got to be presumed we need some sort of small but stable power delivery. May 22 '18 at 15:58
3

If you're worried about the cost of scaling the system to large numbers of livestock, then keep the equipment off the livestock.

  1. Send a drone with a camera overhead periodically. Livestock need to be optically distinguishable from the surrounding ground.
  2. Multiple cameras up on poles. Computer provides triangulation. The more height, the better. You get more frequent updates than #1. This would be easier if the area has roof/fencing instead of an open area.

Are you wanting to track where individuals go, or just to know how many livestock are in a specific area? Counting animals is easier than keeping track of where animal #42 goes.

2

Using RFIDs will be very costly for such large coverage area. It will require a bulkier system. It will be better if you use LoRaWAN. This technology being cheap, low powered, long transmission range capability and durability of more than 10 years will be very helpful in your use case. You can attach LoRaWAN receivers on your livestock and setup the main LoRaWAN transceiver at the center (or corner or even your house if it is close enough to your paddock) of your paddock. You can find a lot of information on Google and a review on this technology on YouTube here.

1
  • 1
    10years ? Where does that comes out from ? Are you saying you can use a GPS + LoraWan in almost real time ? Or that at least you can poll the position whenever you feel like to ? I don't feel like this is true nor a good solution?
    – YCN-
    Jun 21 '18 at 8:19

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.