SoylentNews is people
SoylentNews
When you've got a crop full of plants growing in a field, inspecting each and every one of them can be very monotonous work. That's why scientists are working on plant-inspecting robots, that perform the task autonomously. Most of those 'bots are wheeled, however, meaning that they could get stuck or fall over – plus they might get in the way of other machinery. With that in mind, scientists from Georgia Tech have created a prototype robot that swings over the plants like a monkey. It's called Tarzan.
The idea is that in fields where a Tarzan robot is being used, each row of plants will have a tightly-strung guy wire running overhead. Using its two "arms," the robot will swing itself along that wire, imaging the plants below with its built-in cameras as it does so. When it gets to the end of one row, it will just swing over to the wire running above the next row over, and start making its way back down it. That process will be repeated, until it covers the whole field.
It does sound like a better option than a wheeled robot--muck in the fields can get pretty deep.
(Score: 2) by butthurt on Thursday April 20 2017, @04:20AM
> You don't need detailed analysis of every plant. Crops aren't managed that way. They are managed in bulk.
Broadcast spraying of herbicide, together with GM crops that can resist the herbicide, has served us for millennia. But some people don't want to ingest herbicide, and weeds are becoming resistant. This particular robot doesn't do weeding, but there has been research into robots that will:
https://duckduckgo.com/html/?q=robot%20%22computer%20vision%22%20weeds [duckduckgo.com]
https://duckduckgo.com/html/?q=%22machine%20vision%22%20weed [duckduckgo.com]
A possible use for this particular design is suggested by the article:
[...] without having to spend hours stooped over in the fields, the farmer could find out if any of the plants were showing signs of dehydration, disease or other problems.
Insects, viruses or fungus can be addressed once they have spread to large numbers of plants, but looking at the individual plants would seem to offer the possibility of controlling such outbreaks earlier, before they've done much damage: early detection. If the agriculturalist uses pesticides, less pesticide could be used.
The part about dehydration does sound questionable to me. I would imagine that there could be benefits from irrigating plants individually rather than irrigating an entire field in one go, but it seems unlikely to be worthwhile.