Old MacDonald Had an App

September 16, 2015 • Blog Post • By Atlantic Re:think


  • "Smart farms" are using sensor-driven technology to improve efficiency of farming and harvests, an approach now called precision agriculture
  • Using advanced data analytics, farmers can cross-reference information from sensors to analyze soil trends and weather predictions, making adjustments to their farming practices to better tend their crops

Smart barns, GPS-guided tractors, self-watering fields and self-fertilizing plants: Welcome to the futuristic world of "ag-tech"


Earlier this summer, Jesse Vollmars iPhone buzzed. Something was wrong with his family's corn crop. Vollmar was hours away from the 1,200-acre farm in northern Michigan, so he showed the alert to his father, who had just been there. His dad told him not to worry. The corn had looked perfectly fine from the road.


But Vollmar wasn't convinced. The 26-year-old was using a program he co-developed called FarmLogs, which uses Big Data to monitor fields remotely. So Vollmar jumped into his car and headed north to the plot in Caro, Mich., that had been in his family for five generations.


"Sure enough, once I got through the good-looking corn, and walked out to this spot where my phone was guiding me, there was a big patch of field where the corn was exhibiting this yellow color that was much lighter than the rest of the field", Vollmar says. "Some severe rains had washed away a lot of the nitrogen."


The Vollmars quickly applied more nitrogen fertilizer and salvaged the crop.


"Farmers have always been really good at using their gut instinct to guide their decisions," says Vollmar, who grew up on his familys farm. "But now, we are now so much better connected to our fields, we are able to spend our time way more effectively."


The spread of smart farming


The Vollmar's farm is one of hundreds of thousands of smart farms now using data analytics to improve their yields, an approach known as precision agriculture. According to a recent survey by the American Farm Bureau Federation, nearly 40 percent of respondents are using sensor-driven technology to increase outputs. And as the population is expected to grow by 2 billion before 2050 - calling for 70 percent more food-farmers, and investors, are looking for new, more efficient ways to grow crops.

Rajiv Khosla, a professor of precision agriculture at Colorado State University, says the seeds of the modern smart farm were planted about 25 years ago, when farmers were given access to military-grade technology, like GPS-guided tractors and satellite imaging. Now, GPS systems are so advanced that farm equipment could actually drive itself if laws didnt require operators. The rise of autopilot systems has now allowed farmers to increase their productivity significantly. Crops can be harvested day or night, and farmers working 16-hour days can literally put the machines on cruise control.


"When they are trying to get the field harvested, they can get the work done in a timely manner, and they can do 105 times more work by employing this autonomous technology", says Khosla.


Meanwhile, highly accurate sensors can now be mounted on everything from combines to individual water nozzles. Sensor networks communicate information gathered from the different sensors wirelessly, and farmers can gather detailed views of their fields in real time.


Now, as a combine moves through a field, it can capture the speed, the amount of grain and the moisture content of the harvest. Farmers, in turn, can quantify and predict their outputs in exact detail from every square foot of land. And the machinery can respond accordingly.


"These fertilizer applicators will be smack on the spot where a sensor had been three seconds earlier", says Khosla. "The fertilizer applicators translate that information, and the tractor applies the right amount of fertilizer."


The advent of advanced data analytics means farmers can now also cross-reference information gathered from millions of sensors with existing data setssuch as soil trends and weather predictionsand adjust to the needs of the crops accordingly. More efficient farming practices have also led to more sustainable practices. Many farmers are now able to cut their use of water and nitrogen while still increasing their grain yields.


"This is making agriculture more efficient because we can do more with less", says Khosla, "and it is making agriculture more productive, and it is making agriculture more profitable."


In 2013, farmer David Hula brought in a record-breaking 400-plus bushels of corn per acre using these technologies. But the very next year, Randy Dowdy busted that record with 500 bushels, three times the U.S. average.


"Big Data is a big opportunity in agriculture, but we are just at the tip of the iceberg", says Khosla. "Today we have technology where we can apply differential rate of water from every nozzle level; I can apply how much water where I want, when I want. But how do I know the right amount of water to apply, for the right amount of time? Technology is ahead of science right now."



Deals in the billions


That is slowly changing, however, as more companies, many backed by Silicon Valley investors, hope not only to diagnose the problem but to solve it as well. In 2013, Monsanto bought Climate Corporation, which allows farmers to lock in rates for crops using weather data, and since then, the ag-tech space has been a popular investment. Last year the farm tech industry raised $2.36 billion in 264 deals, and this year Google Chairman Eric Schmidts Innovation Endeavors and Flextronics Lab IX launched Farm2050, a collective that will support ag-tech startups with capital, design, manufacturing, and test farms to try out their inventions.


Innovation Endeavors just invested $9 million dollars in CropX, a mobile app that uses sensors to collect information about topography and soil moisture levels to determine the exact amount of water needed for each field. The company is also backing Blue River Technologies, which has produced a robot to harvest lettuce. Using smart algorithms, Blue River hopes to deliver specific fertilizers and pesticides to individual plants.


"There are a lot of transitions in agriculture", says Dror Berman, Managing Founding Partner of Innovation Endeavors, "and there is definitely a generation change. The next generation of farmers and the industrial farms have taken a much more automated approach to farming, partly it is out of necessity. We will see more and more of that."


Ryan Jacobsen demurs. A farmer and head of the Fresno County Farm Bureau, Jacobsen says the robots wont be taking over his farm anytime soon.


"We talk a lot about technology and automation and how agriculture is evolving to do more and more with less and less, he says. But none of this happens without people. There is no machine for picking a fresh fruit or a tomato. It takes the touch of a human being to run these technological systems."


PODCAST: Machine Learning, AgTech and Tensorflow

Blog Post

Discover Las Vegas: My highlights from the highlight of the year

Blog Post

Intelligent IoT powers Purdues digital agriculture initiative for food security worldwide

Blog Post