Tool innovation sparked the first agricultural revolution 10,000 years ago, but today new digital tools such as robotics, machine learning and data analytics helping farmers feed billions of people around the word.
In a world continuously transformed by digital technology, farming is ripe for significant change.
“Agriculture has been one of the few industries that hasn’t been substantially disrupted,” said Lt. Governor of California Gavin Newsom at the Forbes Ag Tech Conference in Salinas, Calif., located just south of Silicon Valley.
California is the most productive and profitable food producing state in the U.S. Its farmers irrigate nearly 10 million acres of land, which amounts to about 80 percent of the state’s available water, according to data reported by progressive news site ThinkProgress.org. Suffering from five straight years of drought, Newsom urged ag and tech experts to join forces to remake farming into a more efficient, ecologically sustainable industry.
Twentieth century breakthroughs, including geneticist Norman Borlaug’s development of disease-resistant wheat and the combine-harvester, combined to help feed a global population explosion. But with an additional 1 billion more mouths to feed by 2030, computing technologies are critical to the future of farming, particularly ones that collect and analyze data from a long list of interrelated factors like seed, crop inputs like fertilizer and water, weather, farm machinery maintenance, harvest, transportation and food safety procedures.
“Agriculture is an extremely science-based industry,” said Tony Franklin, Intel’s Director of IoT Public Sector and Agribusiness.
He points out one big challenge: identifying the right data to collect and distilling it into simple recommendations. It’s the Holy Grail called precision agriculture.
“Planting at the right time in the right place using the right resources,” Franklin said. There’s an opportunity here for IoT technology and computing power to accelerate the analysis of all of this data coming in. It’s also critical to have the ability to integrate data along the value chain to improve quality and traceability of our food as it moves from farm to fork.”
Sara Menker, the founder and CEO of Gro Intelligence, helps farmers make sense of an extensive collection of agricultural data so they can make informed decisions.
“When a farmer in any part of the world can have access to the data they need, whether that’s soil information or capital markets, that’s transformative,” she said.
Keenan, manufacturer of high-performance feeder wagons for dairy and beef cows works with Intel IOT solutions to automate data sharing and enhance nutrition analysis tools. By figuring out how to keep animals healthier, Keenan is increasing its yields and the tech can help other farmers around the world.
One exhibit at the Ag Tech Summit, showed how big data, artificial intelligence and machine learning combine to form a death-match robot to combat a mortal enemy of most farmers: weeds.
Robovator is a high-tech tractor-mounted robot trained to automatically remove weeds with stealth-like precision.
Weeds can cause crop losses worth $43 billion annually in North America, and that’s just for corn and soybeans, according to a recent study by the Weed Science Society of America. The report suggests that as weeds become herbicide resistant, new technologies will be needed.
From poisoning to physically removing weeds, robotics are showing great promise, especially when it comes to high-value crops such as lettuce, broccoli, strawberries and fruits and vegetables.
“It’s basically computer vision,” said Steve Fennimore, a scientist with the UC Davis Extension in Salinas, describing the Robovator.
“It looks down at the rows with single line cameras, which means each camera sees one line of pixels at a time. It processes them, recognizes the pattern of lettuce, and weeds around it.”
For more than a year, Fennimore and his team have been evaluating the machine which was developed by the Danish company F. Poulsen Engineering.
“We’re finding that it does reduce hand weeding time by about a third,” said Fennimore.
Frank Poulsen, the engineering brains behind the technology, anticipates that computer vision and machine learning will advance the speed and accuracy of robots in the field and revolutionize cultivation.
“Until now computers weren’t widely used in the field, but I think the potential is huge especially in labor intensive areas like vegetable harvesting and sorting,” he said.
Tedious tasks of weeding, harvesting, sorting and packaging can be done with autonomous equipment, according to Carl Vause, CEO of Soft Robotics.
Borne from Harvard research project aimed at emulating an octopus, the Soft Robotics robot uses rubber grippers to gently pick up and sort fruit and vegetables. With two dimensional black and white computer vision, the robot sees shapes and carefully packs items like peaches or tomatoes.
Robots may be a stretch for many farmers, but real-time data that’s accessible on smartphones and tablets are proving to be practical and effective, according to Kevin McCarthy of Davis Instruments, a maker of field sensors for agriculture.
Information about factors like wind, rain or soil temperature helps farmers know when to fumigate or how much to irrigate a field in real time.
“We’re trying to move as much weather data and soil information as we can out from the environment to the farmer on a smartphone or a tablet,” he said. “It used to be a farmer took the temperature and compared it to a chart of averages. Now they have real actionable data about what to do.”
Data is replacing educated guesses. An analysis of crops can done in the morning, and by 3 PM someone is in the field making the adjustments. According to DroneDeploy, an aerial analytics company, it only takes about five hours to stitch together the images from a 37-minute drone flight over a field and have it back to the client.
Using NASA’s normalized difference vegetation index (NDVI) to analyze crops, aerial companies can provide a wealth of information. Besides studying fertilizer distribution or where weeds are growing, a bird’s eye view can save water in critical drought areas like the Salinas Valley. Data driven images showing topography can help farmers understand specific paths of run off in order to better harvest water when it rains.
In a test lot on the campus of Hartnell College, Phil Stiles of SkyViewHD showcases a drone that hovers above a field of newly planted lettuce.
“In being a drone airline,” he exclaimed though a megaphone, “all our pilots are actually pilots. They can fly airplanes, but the computer is doing the heavy lifting.”
The pilot does the take-off and landing, but the routes above the field at about 300 feet are done with a pre-mapped flight pattern.
“We can just sit back and let it go,” he said.
While drones take over the analytics payload for smaller acreages, airplanes still rule the skies for large agricultural operations. Bob Westbrook of TarrAvion said its cameras and technologies can photograph down to 18 centimeters.
“Our next system will be capable of seeing down to 5 centimeters,” he said. That means being able to see each ear of corn.
“It’s really profound, if you think about it, because these aren’t images that get taken and get to the farmer two weeks later. It’s same day. He can go out in the field with a little flashing blue GPS locator on his smartphone and walk right up to the plants.”
Asked what’s in the pipeline for the next technological advance, Westbrook shakes his headsaying it’s all happening so fast he’s stopped making predictions.
“What comes next is coming fairly quickly.”