• Subscribe

Old McDonald had a robot

Speed read
  • Food production faces challenges from population growth and climate change
  • AI could help make farming more productive and cause less environmental harm
  • AIFARMS Institute is developing an autonomous farm of the future to test new technologies

On a farm deep in America’s bread basket, teams of small, low-cost autonomous robots roll through rows of recently planted corn, removing weeds without harming the new shoots—and without the need for herbicides.

Teams of small robots like TerraSentia may one day roam farms, removing weeds, monitoring plant health, and applying targeted fertilizers. Courtesy University of Illinois.

As the corn grows, smart sensors monitor the plants’ health and growth. Another team of robots makes precise applications of just the right amount of fertilizer to boost growth while preventing toxic runoff into local waterways.

Once the corn is harvested, yet more robots seed the field with a cover crop that restores the soil’s nutrients, preparing the field to maximize yield in the next planting.

Growing food in the US doesn’t look like this yet, but it may someday, thanks to the newly created Artificial Intelligence for Future Agricultural Resilience, Management, and Sustainability (AIFARMS) Institute at the University of Illinois.

One of seven National Artificial Intelligence Research Institutes funded by the US government to invest in long-term research that will have a positive impact on societal goals, AIFARMS is tasked with imagining the future of food production. 

“The idea is to develop new ways in which a variety of technologies can be used to make farming more sustainable and more environmentally sensitive—and yet more productive at the same time,” says Vikram Adve, leader of the AIFARMS Institute.

A lot of hungry mouths to feed

While it may seem that wealthy nations in the West have plenty to eat, global agriculture faces a big challenge. “We have to be able to feed a population that’s expected to grow by an extra 2 - 2.5 billion people by 2050,” says Adve. “And we have to do that without more land, and without more labor.”

<strong>The Green Revolution</strong> of the mid-20th century increased global food production, but also resulted in serious environmental impacts such as overuse of chemical fertilizers, water depletion, and soil degradation.Technology is an obvious solution, but today’s farmers don’t want to repeat the mistakes of the past. During the Green Revolution in the mid-20th century, production of cereal crops such as wheat and rice more than tripled. But along with those gains came a new dependence on chemical fertilizers and pesticides, excessive water use, soil degradation, and other serious environmental impacts.  

Adve and the other researchers at AIFARMS hope that a combination of smart sensors, autonomous robots, and data-driven decisions will be able to bring the same or greater gains without the same harmful consequences. To that end, AIFARMS will develop a prototype autonomous “farm of the future” to test experimental technologies.

Down on the farm 

Some of the most important technologies they will investigate are sensors for monitoring soil characteristics such as temperature, moisture levels, nutrients, and interactions between the soil and plant roots. Such technologies do exist, but are expensive and right now can only be used at a small scale. In order to effectively revolutionize agriculture, these devices have to be affordable on a large scale.

<strong>Livestock farms in the US</strong> face a labor shortage. Computer vision paired with intelligent machines could help make up for the shortfall by monitoring and predicting animal behavior. Courtesy US Department of Agriculture.And it’s not just plants—animal farming is also ripe for improvement. Livestock farms in the US currently face a labor shortage, thanks to decreasing availability of migrant labor and a shrinking rural population that is aging faster than the rest of the country. To make up for this shortfall, computer vision could be enlisted to monitor animal behavior. But for this to become a reality, scientists must develop more advanced computer vision and algorithms that can gain a deeper understanding of what is captured on a computer image or video.

But no matter how many hours of dairy cow video footage is streamed from the milking parlor or terabytes of data are gathered from sophisticated soil sensors, the data is of little use without the knowledge to apply it correctly.

Right now, there is no digital substitute for the kind of sophisticated intuition that comes from an experienced farmer. Most current AI algorithms are black boxes with no real understanding of the problems they are called upon to solve. For an autonomous farm to become reality, researchers must find a way to apply specialized knowledge to computer decision-making.

These are ambitious goals, and an essential ingredient in meeting them is computing power—lots of it. Agriculture generates large volumes of very diverse data, from individual plants to entire sectors, from single farms to whole regions, from fractions of a second to seasons and years.

“These kinds of decisions require large analysis and heavy-duty data science,” says Adve. “High-end computing is a key technology to enable all of it.

Training future farmers 

It’s not just technological solutions that AIFARMS hopes to come up with, but sociological ones too. 

<strong>Future farmers.</strong> AIFARMS will be training future farmers to develop more competence in machine learning. But scientists will also have to identify which technologies are most desirable to farmers. Courtesy University of Illinois. “We have a proactive goal of identifying what technologies are likely to be more adaptable than others. And trying to emphasize development of those technologies so that we have a greater chance of having them adopted in practice,” says Adve.           

AIFARMS will invest in training the workforce of the future in a combination of digital and agricultural capabilities. They want future farmers and agricultural businesses to have more competency in machine learning, AI, and other computing skills. But they also want researchers in AI to develop a greater awareness of the scope of applications in agriculture and develop tools that can solve a farmer’s biggest difficulties.

“I have a lot of optimism that technology will help, that it will make a real dent in these problems,” says Adve. “I think the greatest uncertainty lies in how long it will take for that technology to have an impact.”

Read more:

Join the conversation

Do you have story ideas or something to contribute? Let us know!

Copyright © 2020 Science Node ™  |  Privacy Notice  |  Sitemap

Disclaimer: While Science Node ™ does its best to provide complete and up-to-date information, it does not warrant that the information is error-free and disclaims all liability with respect to results from the use of the information.

Republish

We encourage you to republish this article online and in print, it’s free under our creative commons attribution license, but please follow some simple guidelines:
  1. You have to credit our authors.
  2. You have to credit ScienceNode.org — where possible include our logo with a link back to the original article.
  3. You can simply run the first few lines of the article and then add: “Read the full article on ScienceNode.org” containing a link back to the original article.
  4. The easiest way to get the article on your site is to embed the code below.