David Dorhout of Ames, IA, is trying to develop a system where a bunch of robots perform all the farming tasks in a field.
This is a short video that lays out the new concept of swarm farming and demonstrates the first phase with Prospero, the robot farmer.
28 thoughts on “Swarm Farming Explained”
Comments are closed.
ADVERTISEMENT
Support Make:
Join Make: Community Today
Join Make: Community Today
Farming using fertilisers and pesticides is not sustainable. Companion planting, permaculture, seed collecting and saving: all of this can be achieved with far less effort than arranging the planting, artificial fertilising and mechanised maintenance of the crop.
Read up on Permaculture for some ideas of how to get the plants you are growing to help you with the process of growing the plants :)
Hmm couple of big problems with this, which I’m sure they are working on: 1. What if it rains? The UV dye gets washed away. Bot shorts out. 2. Batteries? At the slow rate that these bot swarms work, the equivalent environmental impact compared to a diesel (or biomass!) powered tractor is laughable. 3. Seeds. It takes bushels of seeds to plant a big field. I don’t see any hoppers supplying the seed. How far will the swarm get before running out?
The idea earns cool techie points, but what do the farmers who really know this stuff have to say? All of the phases of large scale farming take the brute and brawn of diesels and heavy iron, according to my limited knowlege.
Aren’t you a bundle of positivity…
It’s an early prototype – The first plane to fly was not the Boeing 747…
If you were to go to any large scale farm today and get into one of the tractors, you would feel as though you are in the cockpit of an airplane. Aren’t these tractors just one large robot? I know for years now my farm and many other farms have used gps and other new technologies to nearly eliminate the need for an operator. When planting a field we are able to use auto-pilot, the tractor is guided by satellite down the rows, keeping everything in line. At the end of a row we still have to turn the tractor around, but then its back to hands off, just watching to make sure everything is going right. With the accuracy at which these seeds are already planted, I find it hard to believe that these swarms would be able to plant with any better spacing. The gps allows you to know what has been planted, so there is no need to check the soil to see if a seed has already been planted there.
I do think it would be cool to see a bunch of robots running around the fields though. I am from ames too, maybe I could find out where these are being tested :)
I was also wondering how this could improve planting. The robots are also so small that any wild animal bigger than a squirrel could take it out. Maybe they could work better on very wet land where a tractor would get stuck otherwise?
The real improvements really need to be made in harvesting. Crops like lettuce are still harvested by hand.
The idea is to bring in the first stages of robotic farming into play. Nothing goes to its best possible version in a day, its a gradual process and what this signifies is that the work on robotic farming is underway and is in the right path with swarm robotics. Realistic robotic farming is nearly a quarter century away, but is definitely there in the future.
Can someone say skynet? ;)
Cool ideas like this will eventually become reality and do all sorts. Hell, I have some crude plans for a couple of wireless lawnmowers which talk to each other.
Just about every commenter on any subject relating to AI or robotics can say skynet. It’s pretty tedious at this point.
It’s an intriguing concept. I would hold some of the same reservations that some other posters have mentioned, in that how is this that different from some of the more high-tech methods used in current large scale farming, but I think the idea is worth exploring.
Off the top of my head, I could think of a few potential advantages. The first would be time; the robots could potentially be operational 24-7, less recharging time. Even if they were slower than the farmer in his modern tractor, they might be able to cover more area over a stretch of several days.
Secondly, not all the robots would need to carry the same seed. As another poster pointed out, a system of companion planting could thus be implemented at varying densities for those who wanted to pursue a more organic path. Personally, if I was developing the idea, I’d likely go with some sort of wireless communication to a central farm server rather than using dye markers to map where the robots planted what.
Another thought is that something like this might be more useful in areas that would otherwise be marginal for agriculture. Steeply sloped areas, small or irregular patches, tightly contoured plots, or out of the way plots that are too difficult to access with large equipment.
Great video! Check out our work at Boulder/MIT on precision agriculture robots that will eventually enable permaculture precision agriculture
http://correll.cs.colorado.edu/?page_id=225
This is very cool, but surely they’re over-hyping the need to eliminate human operators?
Aren’t parts of the midwest USA facing the opposite problem? Entire communities once supported by farming, are now almost entirely redundant thanks to mechanized farming? Not that I’m arguing against progress, but I really don’t think food production is being limited by a lack of humans to drive tractors.
I love the technology but I think this idea will need to be made hundreds of times larger to become efficient. I am just thinking of the number of robots that would be required to manage a few sections of land. The repair and replacement needs would be tremendous. As nickokland mentions there is already a ton of technology in the large tractors that are operated on large farms. I can see the future as improving efficiency of these large machines by adding sensors so it can skip over non-fertile land, only use the required amount of fertilizer, drive an optimized route to ensure fuel savings, etc. I also think with not too much additional work these current machines could be made completely automatic so that no human intervention is needed. The farm of the future could be similar to the military of today where there are a number of unmanned vehicles being controlled and monitored by a single central command post.
I think the flaw in the idea is given during the setup. If there are that many people on earth, then there are that many people who are willing to do this work for much cheaper than developing and building all these robots. It would make sense if there were not enough people to do the farming, but then you don’t need any fancy farming techniques. Maybe if the people of Earth were also supporting a large population someplace not earth. I suppose this technology only makes us ripe as a target to a rapidly expanding alien race looking for “farm planets”.
Food is indeed a basic need and the skills level required to grow it is low, so there are always people who are willing to break their back to perform the necessary tasks.
It’s great that this guy is taking farming seriously and trying to think of fresh solutions for it. I just wish he’d spend a little more time doing his homework, and not plotting fake graphs that rehash crusty Malthusian scenarios about food supply. The biggest challenge in agricultural productivity right now, hands down, is energy. Land quality comes somewhere after that. Manual labor is beyond plentiful.
By the way, why do I have to accept third-party cookies to comment around here now?