Agricultural Robot

I was not sure where to post this so I am guessing this is as good a spot as any other.
I was listening to a podcast recently where the host was discussing the large amounts of money that seed companies and chemical companies are sinking in to research how to get rid of super weeds. A thought struck me that maybe a group of smart people could develop some small robots that could help eliminate some of these super weeds.

I thought that if we could make the robots open source and easily made with say a 3D printer that a farmer could use some of the large amounts of $$ they spend on chemicals on small robots in stead.
So, keeping that in mind I continued my line of thinking and came up with some criteria I thought might make for an interesting exercise.
The robot would need to be
1.) Cheap and easy to reprint/recreate ( 3D Printable)
2.) They could work in groups using Arduino type processors and swarm a field.
3.) They need software to identify the weeds
4.) A simple mechanical way to dispatch the weed (snips or pull)
5.) Rechargeable battery maybe they only work for a few hours and then need a recharge.
6.) Have the ability to navigate a field and return to base (for charging)
7.) They need to be light and easy to deploy.

It would sort of be a Roomba for outdoor use in a field of crops.
I have been debating with my son of a spider sort of design would be better versus something with small tracks.

Now I realize that some of this may be crazy, but UAV's are already doing some of these actions.
I would love to hear what other people think of this. Especially farmers who could give us a good idea on who much $$ they spend on pesticides and herbicides.
I have no idea on what a cost per acre would be. If the cost per acre is low, then maybe this would never make any sense financially.
Also, obviously mono cropping is less than ideal but maybe....just maybe there could be a use for this sort of small machine.

-Kevin

There are prototypes already. Narrow little things that go down one row at a time. Can seed cover crops as well.

Not cheap, and 3D printing is not farm duty yet. Farm spec needs to be a little more rugged than mil spec.

Could you point me in a general direction where I could learn more about these prototypes?

Thanks
-Kevin

Starting with something a bit easier than picking weeds from non-weeds, here's an interesting article on an automatic fruit picking robot.
Vision System


And here's a video on an automatic fruit picker.


It turns out something that seems to be ridiculously simple like picking ripe fruit is actually pretty hard, but humans have evolved to be crazy good at it.

Www.rowbot.com and search on YouTube for rowbot. Not the only one, but the only one I could find on my iPad

Another member of Permies started a thread: NEED IDEAS!!!
This idea of yours would seem to be a good fit.

Agricultural robot, great idea. Weed pulling robot, no, weeds are almost always a symptom of a process error. I used to love double diging, but at 67, not so much. I don't need a robot to identify weeds, I just want a digger tthat knows what pipes and sprinklers are or maybe just one that follows a string.

Great idea when is your first prototype coming out.