4
\$\begingroup\$

My father and I want to develop a cow pedometer to measure a cow's activity. The idea is that if you can measure a cow's activity you can detect when she enters heat. Detecting heat with sensors will help us increase the probability to choose offspring sex depending on insemination time since heat detection, and increase insemination rates since we are not relying on humans to detect cows in heat.

Commercial options are too expensive for us so we have decided to develop the sensors ourselves. I have no EE background, just CS. I come to you for guidance on how to proceed with the project.

Project background

We want to create a pedometer that can be attached to a cow's feet and is powered by a battery. The pedometer must be able to send the accelerometer's data to a central system for data logging and processing. The maximum distance a cow can be from the central system is 1.5km. There are about 400 cows. The farm is on flat ground and doesn't have a clear line of sight (trees block the sight, although it's just a few of them). We also have a mountain overlooking the valley where the farm is (maybe to but up the main receiver?) SEE FARM MAP. Trees in green, farm limits red, circle of radius 1.5 with center in house covers the whole farm. enter image description here

Here are the questions I have, any question you could give feedback on would be great.

How can I achieve a low power long range transmitter? I suppose it will be low bandwidth (only sending sensor id, x accel, y accel, z accel)

Are there any available systems I can use to reduce development time?

At first I will be prototyping with Arduino.

\$\endgroup\$
7
  • 1
    \$\begingroup\$ Your main problem is going to be cost - if you need an accelerometer, RF transmitter, and Arduino for each cow, that's at least $20,000 (excluding batteries, casing, straps, etc, which could easily double that figure) and a hell of a lot of work to assemble them all. If that still sounds ok, this might be a good starting point for the RF side. \$\endgroup\$
    – Polynomial
    Commented May 28, 2015 at 14:37
  • 3
    \$\begingroup\$ Can you put the receiver unit on a drinking trough or some other location the cow will regularly go to? Then you won't need a long range RF link (which will need a big battery and cost money). Use sparkfun or seed to get the parts for a prototype. \$\endgroup\$
    – Jon
    Commented May 28, 2015 at 14:41
  • 5
    \$\begingroup\$ Do you actually need live data? Couldn't you make the pedometer itself do the work of turning that into strides, and only communicate that periodically when the cow is near a receiver? \$\endgroup\$
    – Nick Johnson
    Commented May 28, 2015 at 15:03
  • \$\begingroup\$ @NickJohnson It´s a solution I am contemplating. I need to know how often will a cow pass close to the receiver (when its drinking water, or geting milked) and if that delta time suits for the actions needed for heat detection \$\endgroup\$
    – serpiente
    Commented May 28, 2015 at 15:53
  • \$\begingroup\$ @Jon Thanks for the idea, it´s definetly something I will look into. \$\endgroup\$
    – serpiente
    Commented May 28, 2015 at 15:54

2 Answers 2

Reset to default
1
\$\begingroup\$

According to me using moteino with accelerometer module like MPU6050 would do the trick. One use was able to get a range of 1.5Km using moteinos(https://lowpowerlab.com/forum/moteino/rfm69hw-driverange-test/). You will have to decide the frequency of data that you get, so accordingly the system can go in low power mode and give a longer battery life.

\$\endgroup\$
0
\$\begingroup\$

If you are looking for long range wireless devices, check out these XBees. They can be operated in star, mesh and other generic topologies that allow multiple devices to communicate with each other and can also increase the range in the process.

On how to proceed with the idea, you should follow Nick Johnson's comment. Store the data locally on each individual cow. Place your XBees at strategic places on the field. Scan for cows using their unique ids when they come in the range of your XBee. Check, if this cow's data has already been synced with the central server (placed in your home). If not, download its data and send to the central server. You could also sync data in the night when the cows are sleeping.

Advantages of following this method:

  1. You can implement the data tracking of each cow using low-power devices and BLE. Hence, the battery life on each cow's device will be extended, maybe a week.
  2. You'll need only about 10 XBees all over your field. This would save you a huge cost had you implemented Xbees on each cow.
\$\endgroup\$

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