Farm Data Now Driving Artificial Intelligence

120818-FJR
AgTech Expo ( Farm Journal Media )

Farm Journal held it's second annual AgTech conference in Indianapolis, Indiana. Companies and farmers from across the country were on hand to see the latest, greatest and newest ideas and innovations coming to a farm near you.

These ideas, concepts and companies dotting the expo hall as participants at each booth stand ready to give a pitch.

"Our main goal is to make it as easy as possible for the farmer to use fewer inputs and save money," said Emily Payne the Content Director for a company called Teralytic.

The wireless remote soil probes can sense NPK and several other things in real time.

"It does NPK, soil moisture, salinity, Ph as well as aeration and respiration," says Payne. "One of the main benefits of that is you can see if your traits are leaching down to the 36-inch {deep} sensors and if so, you're losing money by wasting fertilizer."

From tech solutions to innovations in seed science like orange corn from Nutramaize,  the future has many different sides.

"Orange corn has three to five times more {carotenoids} than typical yellow corn," says Nutramaize CEO, Evan Rocheford. "Carotenoids are the same thing that makes yellow corn yellow or make carrots orange."

Then there's the ever-changing flight path of drone technology including new models that include a built-in multispectral camera.

"The multispectral camera is actually capturing pictures on different wavelengths," says Francois-Xavier Charbonnel of Parrot Inc. "You're going to be able to do a map such as an NDVI map."

Even tractors and other equipment are expected to be increasingly autonomous.

"I think for the near future, we're going to automate tractors that we have existing right now," says Tim Norris of Ag Infotech. "I think far off in the future we're going to have a lot of smaller implements."

However, underlying all of this innovation and a whole host of others is data.

"What's changed dramatically the last five years is the introduction of faster processing the mountains of data that are now available to train these algorithms so they are useful for people like farmers in the field," says Josh Henritig, the Senior Director for Artificial Intelligence for Earth and Sustainability at Microsoft.

On the farm its yield monitors, planter sensors, weather stations, pests, irrigation and aerial photography that make up the modern data set. Cultivating results from all of this information have taken time. 

"I think the whole industry promised or way over promised what data would do at the onset," says Norris. "I remember thinking, when we get a yield monitor, we're going to have all the answers to our farm."

Norris says it only created more questions. Today, that's finally changing and companies are now deriving value from that data.

"I feel like the whole thing has been bogged down by a lot of companies trying to extract data from the grower," says Pete Nelson with AgLaunch an ag tech incubator. "In the future, {data} will be monetizable just like eco-credits and carbon credits and a lot of other things that aren't fully monetizable now."

Nelson believes the more the data is used, the more standardized it will get helping to make it more useful across the industry. That data is the base for a quickly emerging field driven by computer learning and artificial intelligence.

"Even more exciting is the potential for AI to look into the future to plan around the weather conditions and resource availability and to make decisions about the types of crops people are going to grow into the future," says Henritig.

Companies like Bayer training artificial intelligence to pick hybrids for the farm.

"We have a couple of breeding programs which are completely in silica right now," says Jim Swanson the Chief Information Officer at the Crop Science Division of Bayer.  "There's not an agronomist as part of it and we're using AI to actually predict how to progress the pipeline."

AI coupled with new data ideas like blockchain, making its way from digital currency to helping improve the traceability of products from farm to store.

"Food and Agriculture was a very siloed fragmented ecosystem and had not gone through any meaningful digitization event," says Phillip Harris of Ripe.IO. "So tasks like traceability, authenticity, measurements of quality were very difficult to achieve because of this mishmash of farmers, producers, growers and distributors that don't talk to one another."

Harris says by putting a blockchain in underneath this system it will enable better communication ultimately helping to solve some of those problems.

"It'll give the farmer the ability to see their crop go through the supply chain in a way that they haven't done in the past," explains Todd Janzen the President of Janzen Agricultural Law."

He sees blockchain making it possible for a consumer to scan a steak and see pictures of the farm that animal came from. 

"There can be some downsides to that too because that means that the processor or the end user can also trace it back to the farm," says Janzen. "If there's a problem with it then it's a bad thing but if there are no problems with it is considered a good thing."

Regardless of the form, venture capital is flooding into this new world of agriculture technology. 

"We're really activating people beyond traditional agriculture," explains Greg Deason of Purdue University. "We're capturing the excitement and energy of people that are in engineering and in data sciences." 

Deason says that is creating a new market for developing sensors, gathering data and creating algorithms for these traditionally non-ag experts. 

"In all cases, farmers are not in this just because it just because it's something that makes a difference," says Deason.  "They ultimately have to do something that's sustainable."

For these ag tech startups that means creating technologies at a price point that makes sense to the bottom line. If the ROI is absent its unlikely to ever be adopted.

If recent years are any indication, innovation will fly forward but what lands remain the biggest question.

Comments