Here Come the Drones: Creating Buzz in the Ag Industry

Share this post:

I have a cousin in Kansas who raises corn and sorghum on a 1400-acre family farm. He calls himself an old-time farmer because he likes to walk his fields to see how the crops are doing. But he sometimes climbs on a powered parachute, which looks like a tricycle with an aircraft engine and a parachute strapped to it, and circles slowly above his fields peering down at the crops. He looks for spots that get too little irrigation or too much.

There’s an easier, less expensive and safer way for farmers to get birds-eye views of their fields: drones. In the past couple of years, these zippy contraptions seem to be appearing everywhere, from next-generation package delivery to unwelcome landings on the White House lawn. The potential commercial applications for drones are seemingly endless–everything from public safety to environmental monitoring–but one of the most promising uses is employing drones equipped with cameras and sensors to help farmers figure out what ails their crops.

The market for drones in agriculture represents more than 80% of the total potential market over the next decade, according to a report from the Association for Unmanned Vehicle Systems International.

A team at IBM Research is developing drone technology aimed at two kinds of farmers–those who want to operate their own drones and those who want to hire others to fly drones on their behalf.  The team doesn’t make the drones themselves, of course. They develop cloud services and mobile apps that help people plan information-gathering flights over fields and enable farmers to interact with organizations offering drone-farming services.

Other research teams at IBM are developing systems for analyzing agricultural data. For instance, a collaboration with Purdue University, which combines drones with sophisticated data analytics, recently won a US Department of Energy grant aimed at improving the yields for sorghum used to produce ethanol.

“A lot of farmers today use paper and pencil to keep track of what’s happening on their farms, and use gut instinct to make decisions,” says Robin Lougee, IBM’s agriculture industry research leader.  “Thanks to these new drone and mobile technologies, a farmer can treat each part of a field as an individual with its own requirements for nutrients and protection from pests.”

There’s a strong parallel between drone farming and personalized medicine. In healthcare, the combination of electronic medical records, genomics and data analytics makes it possible to treat cancer, for instance, for a specific individual–zeroing in on just the right mix of drugs. The same goes with farming. By analyzing crops and fields, farmers can apply the right amount of water, fertilizer and pesticide, and plant the right seeds, for a particular spot. That way they don’t just improve their yields, they can boost their profits–by knowing when they’re wasting fertilizer or water on a part of a field where it’s not needed.

The drone farming project is part of an initiative at IBM Research called Smarter Agriculture, which is exploring the potential for using sensor technologies and data analytics to improve outcomes for farmers, ranchers and others in the ag business. You can think of this initiative as an advance scouting party into a promising new market.

“The world is facing scarcity of agricultural resources, such as water and arable land,” says Chid Apte, Director, Mathematical Sciences & Analytics, at IBM Research. “We can help improve the overall yield for the agriculture industry even while optimizing the use of these scarce resources.”

IBM Research’s Smarter Agriculture initiative fits into a major global trend called the Internet of Things. Thanks to advances in sensing devices, Internet conductivity and data analytics, it’s now possible to gather detailed information and make better-informed decisions about everything from operating utility grids and factories to monitoring the weather to improve responses to storms.

In this case, the focus is on “precision agriculture.” For years, early adopters of technology in the ag industry have been using sophisticated equipment and technology to farm more efficiently–such as GPS to guide tractors, satellite imagery to monitor the vitality of crops and sensors positioned in fields to measure temperature and moisture. Drones get more detailed information than satellite images provide and are less expensive and less vulnerable to being damaged than sensors that are installed in fields.

In addition, advances in data analytics allow people to get much more precise and up-to-the minute readings on what’s happening in their fields, and for agriculture companies to improve everything from seeds and fertilizers to irrigation equipment.

The collaboration between Purdue University and IBM Research demonstrates the potential of combining drones with state-of-the-art data analytics. The goal is to collect accurate phenotypic data about sorghum growing in fields (such as the height of plants, the number of leaves, and the size of the seed heads), combine it with genomic data, and use that information to create new hybrid varieties of sorghum for the biofuels industry.

The IBM Research technology analyzes the images of plants captured via drones and tractors, analyzes genomic data, and discovers previously unknown links between the two types of data. It combines machine learning techniques and algorithms. “The goal is to enable plant scientists to breed new and better hybrids much faster,” says Naoki Abe, the lead IBM Research scientist on the project.

These are early days for drones and data analytics in agriculture, but the major ag companies and many farmers are excited about the possibilities. That’s what hooked IBM Research, as well.  “We started the drone farming project because we thought that drones were just really cool technology,” says Justin Weisz, Senior Manager, Mobile Solutions and Infrastructure, who is the drone farming project leader. Now, he and his team are committed to developing apps that are practical and impactful for farmers.

The apps are built on IBM’s MobileFirst for iOS platform, so they run on Apple iPads and iPhones. The team is primarily focused on improving the quality of human-drone interactions. For instance, the apps make it simple for a farmer to survey a field using a drone using sophisticated analytics that plot the most efficient flight path.

IBM has a history of transforming industries with technology, so today’s experiments in the fields and plant science labs could eventually turn into something big–both for IBM and for the ag industry.