Help to Build a Smart Fence to Track Elephant

These elephants, are they captive or free? That is to ask, can the animals be fitted with a tag such as an active RFID?

As I consider further, RFID is probably not a workable idea anyway. Please tell us more about the project. How many sensors, distances involved, operational environment... those sorts of things.

I hope the OP is not an Elephant poacher.......such a device would make his job easier to hunt down an Elephant.

WHY IS SUCH A DEVICE NEEDED???? WHAT IS IT'S PURPOSE???????

This time I'm with Moosie.

Tell us more, or we'll asume that you are up to no good.

My task is to inform the farms of wild elephants. The project is based on sensors so i can't tag them. I have just started my project & haven't thought about the height of the poles & distance between them but they are intended to be used along the boundary of an animal reserve.

Is this some sort of "ADVANCED WARNING SYSTEM" of sorts against rampaging elephants to prevent casualties and destruction of property?

Make a radar !

<sigh>

Monitored Boundaries: Total of 1,321 kilometers, more or less?
Location: Mandalay Burma?
Distance from home station to furthest sensor: 560 kilometers, more or less?
Reason I need to be alerted when elephants are close to the boundary: I like to take their pictures, and I collect their fresh dung to sell to tourists.

You have to give us something to work with. As has been pointed out, this would be a very handy system for an elephant poacher. I am already wary of this: If you cannot provide some useful information, I will join the others in the belief this is an undercover operation, one with criminal intent.

To all above;

Cultural update for the week.

In Asia there is a widespread problem with wild elephants destroying crops outside reserves. There are many types of fence including hand made twine soaked in scent and/or irritants.

A quick google should bring you all up to speed on problems and solutions to date.

It's an every farmer for themselves situation, particularly in Burma, the government being largely useless.

Elephants are also held as 'sacred' in most of Asia, so the idea is to keep them out of the vegi patch, not to harm, kill (or poach) them.

The OP looks like wanting an intruder alarm that sends image to a cell phone (confirming presence) before others are called to chase off the thieves.

What you say is true.

It is also true that Kandy (Sri Lanka) is a center of black market ivory trade. This ivory is from elephants taken from (killed) the herds that roam the country.

While most local indigenous peoples hold the elephant as sacred, the country may have non-local indigenous people who hold no such love of the animal, those people holding more dearly the money that can be had.

I found this which is I think is a fair overview.

Whether the 'black Ivory' is 'poached', or from 'casualties of war' I wouldn't know. I've never been to Sri Lanka.

The question might be; how long until the elephants figure out how to disable the cameras.

Well, a fair overview or not, it is a heart rending story. I don't doubt the story's authenticity, nor do I question the motivation of the author. I have heard similar stories, but your link is very detailed and better written than most.

Wow. Just plain, old fashioned, wow.

From the linked story, I am not sure just what effect a remotely monitored picket line described in the OP will have. It sounds like the battle is lost when the elephants show up.

Thanks 34.5, good general info. Now I have one more thing to be depressed about.

I got this project last week & we didn't pic a location yet but definitely the distance from home station to the furthest sensor will be less than 560km.

I can think of only one method that would be useful if no RFID tag could be applied to the beasts. While I too am slightly suspicious as to the motives behind this, I also think having a permanently installed EXPENSIVE system like like this almost precludes a nefarious intent. Let's face it, poachers are not the brightest minds on the planet and a system that needs routine maintenance is not something an outlaw would be likely to invest in. It's existence alone would be a beacon to their activity and potential trap for law enforcement (if such exists in that area) when they show up to maintain it.

About the only reliable method of detecting a large animal like that is going to be infra-red (IR) cameras. With the right software you can scale an image detector to alert on overall hot image area, that way you can ignore things that are smaller. Laser scanners would not be able to discriminate a change in vegitation shape due to wind from the addition of a large animal, although it may be useful in aiming the IR camera so that it can look for IR content in a changed scan area and avoid constant back-and-forth scanning of the camera.

So, assuming a solar panel and battery pack to run this, an IR camera, a laser scanner and a hard wired telemetry system to send the data back to a central location, figure about US$10,000 per sensor, plus around US$50,000 for a base station with camera multiplexers, IR image scanning and interpretation, video recording, alarm annunciation, cell phone dialer etc.

I'm doing this as my final year project so money is an issue. Higher the value of the equipments higher the probability of them being stolen.

What about installing a parallax PING ultrasonic sensor?

If you use an IR sensor that is mounted high up, that would eliminate dogs but not birds or tree dwelling animals.

If it's an IR camera, the user can recognize elephant vs bird or other smaller animal. Still you have the problem of a camera's orientation and range, and the fact that it takes more power for a camera than it does for a sensor. A lot more afaik.

What about chemical sensors... are elephants smelly? That wouldn't take much energy for the sensor device.

Actually come to think of it, the weight of the elephant is very much a defining factor. All you really need then is a vibration sensor in the ground, which would tell you for sure that heavy feet are trampling in the area. These don't take a lot of energy so it's practical to do with solar.

Ooooh, I like the vibration sensor idea, hadn't though of that. It wouldn't be able to discriminate between a truck and an elephant, but for the most part you would likely already know if a truck was in the area.

Might be an issue in a seismically active area though.

There is no cheap defense from a hungry elephant in this case. Having a sensor network will offer nothing but a brief warning of imminent destruction of life and property.

You'll have to devise a way to convince humans to "un-encroach" from the elephant's habitat.

That ain't gonna happen.

Apparently they're doing really well with the chili fences, getting the elephants to 'unencroach' on the farmers' habitat.

Lights work fairly well also, to send them scurrying off, and loud noise is a classic. So a sensor system would be useful, to tell people when to turn those on or, hell, just wire it up automatically to turn on the light show. Wake the farmers with a loud siren, or such, so they know the elephants are being frightened.

_{I need a moose sensor system.}