Go Down

Topic: Long Range (~200ft) Aerial Tracking System (Read 458 times) previous topic - next topic

Jgunde

Hello everyone. I would like your advice on this project:

I am trying to make one RC plane follow another RC plane autonomously. I having everything figured out except one problem: How do I have one plane track another outdoors with a range of about 50-200ft? (*GPS is not an option - I want a line of sight based tracking method)

Possible Ideas:

- use bright IR led on leading plane and IR receivers on the tracking plane: the problem I see with this is the IR led not being bright enough to be detected by the IR receivers in daylight

- use ultrasonic emiter on the leading plane and multiple receivers on the tracking plane: again, I don't think the ultrasonic pulses would be able to travel up to 200ft

- use RF transmitter on the leading plane and multiple RF receivers on the trailing plane; the receivers would compare the time delay between when each receives a RF signal to determine the direction of the leading plane: This seems like the most plausible solution but I would imagine hardware that can differentiate time differences in the sub-nanosecond range would be expensive (though LIDAR units are becoming cheaper) and I have no idea how to do something like this.


If anyone has any better ideas or suggestions on how to make this work, all advice would be approciated.

couka

#1
Jan 12, 2019, 09:55 am Last Edit: Jan 12, 2019, 10:20 am by couka
I think IR is the easiest to do, you just need a powerful enough transmitter.

This might be interesting (Using the IR camera from a Wii controller)

For the RF method, look into "Amateur Radio Direction Finding"
youtube.com/DieBastler1234
Don't send me technical questions via PM. They will be deleted unanswered.

Robin2

If the planes are 200ft apart the circumference of the circle in which the following plane must find the other is about 1200 ft. If the planes are about 6ft wingspan that represents half of 1% of the circumference. Imagine the tiny percentage when you take 3 dimensions into account, especially as the airplane is not 6ft tall as well as 6ft wide. Acquiring the target in the first instance will be a mammoth task.

Short of using proper radar (and even that might be impractical with such a small target) the only thing that seems to me likely to have a realistic chance of working is a visual system with image recognition software and, perhaps, a flashing visible light on target plane. The flashing is to make it identifiable and to distinguish it from the sun as much as possible. Flying on an overcast day might be wise so that the sun cannot interfere.

You would need a lot more computing power than an Arduino for a visual system.

...R
Two or three hours spent thinking and reading documentation solves most programming problems.

PaulRB

Quote
GPS is not an option - I want a line of sight based tracking method
Why? GPS is an obvious candidate solution, and you don't have many other practical ones yet...

srnet

Is the objective of the 'project' just to come up with an alternative to a GPS solution, or is there some other compelling reason why you cannot use GPS ?
$50SAT is now Silent (but probably still running)
http://www.50dollarsat.info/
http://www.loratracker.uk/

Jgunde

#5
Jan 12, 2019, 04:52 pm Last Edit: Jan 12, 2019, 05:00 pm by Jgunde
Is the objective of the 'project' just to come up with an alternative to a GPS solution, or is there some other compelling reason why you cannot use GPS ?
Why? GPS is an obvious candidate solution, and you don't have many other practical ones yet...
I wanted to replicate the style of tracking they use in the AIM-9 Sidewinder missile. It tracks the IR radiation coming from the enemy jets engine.

So I want a system that can visualy lock on to a target and then follow it.

Robin2

I wanted to replicate the style of tracking they use in the AIM-9 Sidewinder missile. It tracks the IR radiation coming from the enemy jets engine.
There's your solution. Put a 30,000 lb thrust jet on the target plane.

Or, put another way, some things don't scale down.

...R
Two or three hours spent thinking and reading documentation solves most programming problems.

MorganS

This will scale nicely. The sidewinder can track because the hundreds-of-degrees IR radiation does not exist in nature. At those wavelengths the rest of the world is black (with the exception of the sun).

IR remote controls use a similar idea: they transmit in a range that is mostly black in the world.  The other thing that is different is the IR diode is modulated at 38KHz. The world is completely black to the 38KHz filter.  Your TV can see the remote in direct sunshine.

Initial acquisition might be difficult but once you are locked on it should not be difficult to track.
"The problem is in the code you didn't post."

jremington

#8
Jan 12, 2019, 07:18 pm Last Edit: Jan 12, 2019, 07:19 pm by jremington
Quote
IR remote controls use a similar idea: they transmit in a range that is mostly black in the world.
During the day, the world is saturated with NIR radiation. It is the 38kHz modulation that makes an IR remote stand out.

Robin2

This will scale nicely.
From a distance of 200 ft? I reckon it's like a needle in a haystack.

...R
Two or three hours spent thinking and reading documentation solves most programming problems.

Paul_KD7HB

I wanted to replicate the style of tracking they use in the AIM-9 Sidewinder missile. It tracks the IR radiation coming from the enemy jets engine.

So I want a system that can visualy lock on to a target and then follow it.
Just a minor point: long before the missile is launched, it is given the image with the point to track. So, it only has to keep the hot spot centered in the image.

Paul

PaulRB

it only has to keep the hot spot centered in the image.
Actually, they don't. Ground/air-to-air missiles don't aim directly that the target. They attempt to maintain the same angle to the target. This means they are in effect aiming at a point in front of the target, where the missile will meet the target (and of course missiles almost never actually hit their target, they just explode close enough that the chains in the warhead can damage the target).

Robin2

they just explode close enough that the chains in the warhead can damage the target).
And a lot of them miss completely.

...R
Two or three hours spent thinking and reading documentation solves most programming problems.

MorganS

Just a minor point: long before the missile is launched, it is given the image with the point to track. So, it only has to keep the hot spot centered in the image.

Paul
That applies to missiles launched from internal bays like on the F22. For older planes with the missiles on the wings the missile does all the work. The pilot was supposed to listen for the "growl" generated by the missile seeker and launch when it changed to a continuous tone, indicating a lock.

For certain kinds of missile the pilot would always keep one back because if he fired it he lost that type of sensor which the aircraft did not have on its own.

There is a surprising amount of data sent to the aircraft from the missile before launch.
"The problem is in the code you didn't post."

Robin2

#14
Jan 13, 2019, 08:52 am Last Edit: Jan 13, 2019, 08:53 am by Robin2
I suspect that heat-seeking missiles had ultra-cold detectors as well as everything else - not just a regular TSOP4838 that is designed to be as omni-directional as possible  :)

...R
Two or three hours spent thinking and reading documentation solves most programming problems.

Go Up