PDA

View Full Version : [Project] Prospero: Robotic Farmer



Vanmunch
02-13-2011, 12:07 AM
Edit: 10-19-2011

I've put together a short video that lays out the argument for and explains my concept of using swarms of robots for farming. Let me know what you think!

New Video: Future of Farming: Prospero Robot Farmer
http://www.youtube.com/watch?v=CReaedEF41w

http://www.youtube.com/watch?v=CReaedEF41w

--------


http://forums.trossenrobotics.com/gallery/files/5/0/1/6/planter_front_c.jpg http://forums.trossenrobotics.com/gallery/files/5/0/1/6/prospero_walking_c.jpg



http://www.youtube.com/watch?v=ACtihXjq2B0
Overview with extended planting and annotations (5:04)


http://www.youtube.com/watch?v=jdEdV-ct1HM
Just planting (0:43)



Prospero is the working prototype of an Autonomous Micro Planter (AMP) that uses a combination of swarm and game theory and is the first of four steps. It is meant to be deployed as a group or "swarm". The other three steps involve autonomous robots that tend the crops, harvest them, and finally one robot that can plant, tend, and harvest--autonomously transitioning from one phase to another.

Prospero is controlled with a Parallax Propeller chip mounted on a Schmart Board (Schmartboard 2010 MCU contest http://forum.schmartboard.com/index.php/topic,871.msg1038.html#msg1038). Its body is designed by Lynxmotion and the original programming allows it to walk autonomously in any direction while avoiding objects with its duel ultrasonic Ping))) without turning it's body. An under body sensor array allows the robot to know if a seed has been planted in the area at the optimal spacing and depth. Prospero can then dig a hole, plant a seed in the hole, cover the seed with soil, and apply any pre-emergence fertilizers and/or herbicides along with the marking agent. Prospero can then talk to other robots in the immediate proximity that it needs help planting in that area or that this area has been planted and to move on via IR (currently represented with a green and red LED ). The more seeds it plants, the more the "green" LED lights up, the more it draws other robots nearby (+2). The more it detects planted seeds, the more it repulses other robots with the "red" LED (-1)

Why small, autonomous robots?
Robotics in Agriculture
Despite its quaint reputation, agriculture has always been an early adapter of technology. This is evident from the beginning of mechanization with the cotton gin, McCormick's Reaper, tractors, hybrid seed, to genetically engineered plants that protect themselves and grow in arid environments. Yields have grown quickly, but demand from developing countries and population growth are growing faster

We know that we need to continue to find ways to increase the productivity of land on a per unit basis. Agriculture has started to add computerization and automation to the current machinery with things like GPS based precision farming systems that can autonomously drive tractors, monitor yield, and apply fertilizer. However, these aftermarket add-ons are built around the single most expensive and awkward part of the equipment. The person controlling the tractor.

Today's agricultural equipment has been designed around a person sitting in a chair. It cost a lot to employ a single person so the equipment grew larger in order to maximize the productivity of that one person. However, this method has its drawbacks. Farming decisions have to be made at the field level. Nature is chaotic and dynamic. Soil nutrients and moisture change from foot to foot. Having equipment that allows a single person to plant a thousand acres in a day comes at the cost of productivity per acre as a result of treating all those acres as the same. A swarm of small robots like Prospero would have the ability to farm inch by inch, examining the soil before planting each seed and choosing the best variety for that spot. This would maximizing the productivity of each acre, allow less land to be converted to farm land, feed more people, and provide a higher standard of living for those people because they would spend less of their money on food.



Please let me know if you have any questions or if you want me to post the code. Thanks!

Dave

Vanmunch
02-14-2011, 09:58 PM
Here is the project write-up for the Schmart Board contest including BOM and source code

Dave
"vanmunch"

Chench
06-10-2011, 05:25 AM
Very impressive! I like how it is designed to do something out in the real (and messy!) world. Have you set in loose in a garden or field yet? Also the systems for marking the ground and planting the seeds are very innovative. Thanks for sharing!

Vanmunch
06-10-2011, 11:15 PM
Hey Chech, Thanks. I like your scratch built hexapod. It looks great!

I've shot some new video and pictures of Prospero farming outside in a normal corn filed here in Iowa. I'm working on editing the video and should have it uploaded in about 3 more weeks (I'm getting a new project ready for the Unofficial Propeller Expo Central [UPEC] on June 25 near Chicago which has been keeping me busy). There's also going to be an article about it in the next "Robot" magazine and I was going to wait with posting any new pictures until then, but there's a really cool one that I should be able to post in the next day or so as a teaser. :)

Gertlex
06-14-2011, 04:52 PM
Very nice construction there!

Couple questions via my curiosity:
Is the paint going from the can to the ground via a tube?
How close do you reckon the bot is to it's payload limit?

Vanmunch
06-14-2011, 06:43 PM
Hi Gertlex,

Thanks for the compliment and the questions. You're correct, the paint goes through the tube and is sprayed onto the ground. The tube is about 3 inches above the ground. You can see it in the "underneath" picture and the actual spray in the video. I'm just using white primer paint because it's easy to see what's going on, but if you were doing this for real, you could make a custom mixture with fertilizer and a pre-emergence herbicide and use an IR reflective. (no one really wants a poke-a-dot farm field)

The robot itself is pretty much at its maximum payload capacity (~10lbs including the body). It could hold more if I spent more time on the gait which I'm going to do a complete overhaul on once I'm done with a current project that I'm getting ready for the Unofficial Propeller Expo Central in two weeks. It's going to be just south of Chicago and I'm bringing Prospero. I highly recommend it.

Dave

Vanmunch
10-20-2011, 04:01 AM
I've put together a short video that lays out the argument for and explains my concept of using swarms of robots for farming. Let me know what you think!

New Video: Future of Farming: Prospero Robot Farmer
http://www.youtube.com/watch?v=CReaedEF41w

http://www.youtube.com/watch?v=CReaedEF41w

3546 3547 3548

LisaJensen
11-25-2011, 06:27 PM
Hi, I'm impressed with your work. This is exactly the sort of thing I have been looking out for. Since you're interested in increasing productivity per acre as well as per farm worker, I was wondering if you've thought much about companion planting and how you might modify the marking and swarming concepts. I'm thinking a good starting example for you would be the corn, beans, squash triad since you're already doing corn.

Vanmunch
11-27-2011, 10:08 PM
Hi Nelly, LisaJensen,

Thanks for the encouragement. Intercropping is one of the big ideas that could happen practically with swarm robotics. I had wanted to put it into the video, but it ended up being dropped in the interest of time. I'm excited that you're thinking the same thing. Current farmers often rotate between corn and soybeans because the soybeans are able to fix nitrogen via a symbiotic relationship with a species of bacterium from the atmosphere (N2) into ammonia (NH3) that plants can use. The soybeans fix more nitrogen than what they need which then remains in the soil and is used by the corn plants the next year. This soy/corn rotation allows the farmer to spend about 30% less on ammonia fertilizer for the corn than if they were planting corn year over year. The main drawback of a corn/soy rotation is that a farmer typically makes less money growing soybeans compared to corn on the same acre of land. (There are other benefits to rotation, but they are related more to pest and disease management.) Corn and Soybeans are never currently grown together because there isn't a practical way to harvest them separately. Cover crops would have the benifit of water conservation. Having robots that are able to farm inch by inch and plant the best varieties within a species and mix crop species for synergistic effects is going to change farming as much as switching to swarm robots and will be exciting to watch.

Are you involved with food production?

Dave

LisaJensen
11-29-2011, 05:15 PM
Hi Dave,

At this point I'm just trying to automate my backyard food production. I only have 800 square metres to work with. I've already planted most of the perennial food forest, but haven't started the annual veges. I have 5 beds planned, and have diagrams of how to interplant all the companions but I'm not looking forward to the tedium of planting, year after year. I'm also just getting started with robotics and am planning on looking at the harvesting issues in the food forest. Hoping my remote sensing background will help me out. I think navigating around the forest to find the things that need picking will be quite a fun challenge.

Lisa

Vanmunch
10-08-2012, 09:08 PM
Hi Dave,

At this point I'm just trying to automate my backyard food production. I only have 800 square metres to work with. I've already planted most of the perennial food forest, but haven't started the annual veges. I have 5 beds planned, and have diagrams of how to interplant all the companions but I'm not looking forward to the tedium of planting, year after year. I'm also just getting started with robotics and am planning on looking at the harvesting issues in the food forest. Hoping my remote sensing background will help me out. I think navigating around the forest to find the things that need picking will be quite a fun challenge.

Lisa

Hey Lisa,

Any update on your robot?

Dave