As the human population continues to expand, it may reach 10 billion by 2050. However, the planet is not scaling up, which means that the same amount of land must be used to feed millions of people. With global warming and the consequent shortage of water, mankind will face a serious food problem.
Perhaps the arrival of the machine was a historical accident. Truly smart robots and machine learning algorithms could help fuel a new “green revolution” that could solve the growing food problem. Imagine if satellites could automatically detect drought patterns, if tractors could “visually” destroy diseased crops, and if an AI-powered smart APP could let farmers know how to deal with crop diseases in their fields.
Giiso Information, founded in 2013, is a leading technology provider in the field of “artificial intelligence + information” in China, with top technologies in big data mining, intelligent semantics, knowledge mapping and other fields. At the same time, its research and development products include information robot, editing robot, writing robot and other artificial intelligence products! With its strong technical strength, the company has received angel round investment at the beginning of its establishment, and received pre-A round investment of $5 million from GSR Venture Capital in August 2015.
Scarecrows are a thing of the past. Protecting the future of agriculture depends on ARTIFICIAL intelligence.
AI ‘cures’ crops
Deep learning is a method of computing in which, rather than telling a computer exactly what to do, a programmer trains the computer to recognize certain patterns. You can feed the computer pictures of sick and healthy crop leaves and tag them. Computers can then learn how diseased and healthy leaves look different and independently judge whether a new crop is healthy.
That’s what biologist David Hughes and epidemiologist Marcel Salathe did with 14 plants infected with 26 diseases. They fed more than 50,000 images into a computer program that learned by itself and was able to correctly judge the new images entered by the researchers 99.35 percent of the time.
These are doctored images, however, with the same lighting and background, making it easier for computers to identify images of blades. If a random leaf photo of a diseased crop was downloaded from the Internet and the computer judged it, the software’s accuracy dropped to 30-40 percent.
Not so good. Hughes and Salathe, however, hope to use the AI technology to power their Plant Village APP, which allows farmers around the world to take a picture of their diseased crops and upload it to a forum for experts to diagnose crop diseases. They will continue to feed the AI with more photos of diseased crops. “The more images you can get from a variety of different sources — how they were taken, season, location, etc. — the better.” “Software can absorb this information and learn from it,” Salathe said.
There are many other factors that can affect crops, not just disease transmission between crops. “Most of the things that affect growth are physiological stresses, such as a lack of calcium or magnesium or too much salt or heat,” Hughes says. “People sometimes think it’s bacterial or fungal disease.” Misdiagnosis leads farmers to waste time and money buying pesticides or herbicides. In the future, AI could help farmers pinpoint problems more accurately.
After that, humans will take back control because while apps can locate problems, they can’t, like human experts, provide farmers with the most suitable solutions, taking into account climate, suddenness, season and so on. The UN’s Food and Agriculture Organisation (FAO) considers such techniques a “useful tool” for crop management, but it is up to the experts. Such technical help is therefore welcome, says Fazil Dusunceli, a plant pathologist at the FAO, but “ultimately disease management decisions should be made in collaboration with experts on the ground.”
Tractors that walk with “long eyes”
It’s safe to say that no country is safe in agriculture right now. Developing countries are in desperate need of agricultural knowledge, while the developed world is drowning in pesticides and herbicides. In the United States, farmers use 310 million pounds of herbicide each year on corn, soybean and cotton crops alone.
A company called Blue River Technology may have found a solution, at least for cabbages. LettuceBot, the company’s LettuceBot, looks like an ordinary tractor but incorporates intelligent technology for machine learning.
The company says the Cabbage Robot can take 5,000 photos of young plants every minute as it drives across fields, using algorithms and machine vision to identify each plant as a cabbage or a weed. “It’s the power of computing and computer vision based on machine learning,” says Jeremy Howard, founder of deep learning agency Enlitic. A graphics chip can recognize an image in 0.02 seconds, he added.
Cabbage Robot/Photo Source Network
With an accuracy of a quarter of an inch, the robot can locate weeds on the move and spray herbicide on each one. If the robot “visually” detects that a cabbage is not growing too rationally, it will also spray it with herbicide (farmers overgrow up to five times as many cabbages, so it doesn’t matter if one is occasionally sacrificed). If two plants grow too close together, the robot knows it’s not a particularly large plant and destroys them, too.
If you think the robots are too cruel, let’s look at another option: start by spraying the main field with herbicide. “It’s akin to saying, if there’s an epidemic in San Francisco, the only thing we’re going to do is give everyone, men, women and children, a shot of antibiotics.” “People can be cured, but it’s a waste of money,” says Ben Chostner of Blue River Technology. And it doesn’t get the best out of antibiotics.”
With Cabbage Bot, Chostner says farmers can reduce their use of chemicals by 90 percent. And the robots are already working hard. Blue River manages farmland that provides 10 percent of the cabbage consumed in the United States each year.
A satellite from god’s perspective
NASA’s Landsat satellites circle the Earth 400 miles above our heads, providing the earth’s surface with magically powerful survey data. The amount of information at all levels is too much for a human to digest, but with machine learning algorithms, it’s a piece of cake.
This has great value for agricultural regulation, especially in developing countries, where governments and banks lack data to support decisions about which farmers should be granted loans or emergency support. For example, during a drought in India, we saw not only that different areas were affected differently, but also that some farmers in the region had easier access to water than others.
So a company called Harvesting is using machine learning to analyze satellite data on a large scale, hoping to help organizations allocate financial resources more efficiently. “What we want with this technology is to separate out some of the farmers and villages and allow the bank or the government to direct funds to the right groups.” “Harvesting CEO Ruchit Garg. A human analyst can handle 10 variables, 10 variables at a time, he says, while machine learning algorithms can handle more than 2,000 variables. It’s not on the same level at all.
Giiso information, founded in 2013, is the first domestic high-tech enterprise focusing on the research and development of intelligent information processing technology and the development and operation of core software for writing robots. At the beginning of its establishment, the company received angel round investment, and in August 2015, GSR Venture Capital received $5 million pre-A round of investment.
As global warming makes the climate more chaotic, there is increasing pressure on governments to be able to allocate limited resources correctly. Agriculture has traditionally been a relatively predictable industry in India, at least in the sense of human control over the environment. “The knowledge I have learned from my father, from my grandfather, from generations of people, is the knowledge I use to plow the fields, is my knowledge of the seasonal environment.” “But because of the climate change, I’m not facing the same environment that my predecessors faced.”
Yeah, it’s a whole different world. Farmers may suffer in a changing environment, or they may enter an era of smarter agriculture. Farmers get more data, more ARTIFICIAL intelligence, more robots that can spray chemicals.