Get Ready for Generative AI

Ever Heard of Generative AI? It's Coming to a Farm Near You

Dan Miller
By  Dan Miller , Progressive Farmer Senior Editor
AGCO's Senior VP Seth Crawford clues us in on the new field of generative AI and updates us on AGCO's autonomy-by-2030 goal. (Photo courtesy of AGCO Corp)

HOUSTON (DTN) -- Seth Crawford is senior vice president and general manager of Precision Ag and Digital at AGCO Corporation in Duluth, Georgia. Wearing another hat, he also keeps a hand in his family's corn and soybean farm back in Rock County, Minnesota, which was established in 1872 -- 14 years before Progressive Farmer was born.

Due to his professional responsibilities at AGCO, Crawford lives in that intersection where technology platforms written by coders are tested against expectations and harsh field environments common to farming. Do technologies make farmers more efficient and more productive, add value and make farms more sustainable?

"I have a vision," Crawford once told DTNPF, "That in my lifetime, we eliminate unplanned downtime by way of (quickly deciphering) the thousands and thousands of data points we collect going through the field. Because unplanned downtime is super frustrating to the farmer. It can create a snowball effect of things going wrong."

And so, we spoke again Thursday with Crawford, this time during the 2024 Commodity Classic where a record number of farmers and exhibitors -- 10,856 attendees at last count -- would see AGCO's portfolio displayed down machinery row in Houston's George R. Brown Convention Center. Tractors, combines, planters, sprayers, haying tools -- all delivering full-season "solutions" by way of AGCO's Fendt and Massey Ferguson brands with its Precision Planting arm in the center of the display. Not yet on display at Commodity Classic, but coming soon as final papers are signed, are the fruits of AGCO's $2 billion joint venture with Trimble.

DTNPF Q&A WITH AGCO'S SETH CRAWFORD:

P[L1] D[0x0] M[300x250] OOP[F] ADUNIT[] T[]

DTNPF: We all hear pronouncements about AI, artificial intelligence, machine learning, other technologies and approaches. What would you suggest farmers take away from all this?

Crawford: It's one of the most exciting times I've seen, what's possible. The artificial intelligence out there today, it can provide you with a specific how-to. What do you want to do today? Plant corn? Plant in certain conditions? What do I need to be looking for? (That planning process) can be more tailor-made than ever before by way of AI -- when you get data coming off the fields, being able to analyze [and] show it in different ways and (incorporate) different data layers. AI is not new but what is new is that we can take these large data sets, take the computing power available and get to an answer so much faster. You're getting the soil and your farm ready to maximize its output in a much more granular way than you were able to before. Then you get into generative AI, not only a year old ...

DTNPF: Apologies for my interruption, but "generative AI?"

Crawford: Generative AI is essentially predicting the next best (action)). If I'm taking a sentence with the word "and," generative AI would produce the next words (based on relevant information, discussions and other resources). It's just going to keep on generating (solutions) based on everything it has analyzed. I've spent a fair amount of my time in the customer support area. We deliver service to customers to improve their performance. We were able to apply a generative AI tool to service calls. A farmer calls in about his combine; we ask about the equipment on the machine, the conditions of the crop and generative AI is listening in the background. (It is) pulling in the combine manual, information about the front-end equipment, crop types and where we've seen this problem before. Generative AI is giving prompts, questions to ask customers. It's taking a problem that might have taken two weeks to solve, it is giving the service reps questions -- ask the customer about this, ask the customer about that. It narrows in on the problems with all its data sets (in hours).

DTNPF: Tell us about AGCO's goal of having a fully autonomous suite of machinery by 2030.

Crawford: Last summer we were in the field demonstrating the autonomous grain cart. We have autonomous tillage. And, we are automating more and more features (on AGCO machinery). You'll see it when you (look) at Precision Planting, being able to monitor operations, whether it's planting, whether it's air seeding, your sprayer -- the applications of the cameras on the sprayer. Cameras used more and more in the future, (to see) are we getting the job right. We feel we are in a good position to deliver on that commitment for 2030.

DTNPF: The approach then is if you can capture with automation all the actions of machinery inside a cab, on displays for example, you should soon be able to monitor all the activity remotely, away from the actual machine?

Crawford: Absolutely. Today this adds value for the farmer -- it's an aid to the operator to ensure the job is being done right, and it adds incremental value farmers are willing to pay for. But as we go down the road a couple of years, the farmer will say, "You know, I was watching all this on my screen anyway, as I was riding along. It's not going to be any different when I'm completely removed from the machine. If I need that certainty, I can look on my (smart device)."

DTNPF: What do farmers tell you about autonomy?

Crawford: It's highly mixed. Some say that's the place they want to be in the cab. It's their "me" time. Others say finding labor for three or four weeks of the year -- or putting myself in that seat for 16-hour days -- it's just too much. I think what scares away (some farmers) is the uncertainty. We show this does work, you need to trust this, and in fact, it does the job better in a consistent manner. With that, adoption will take off. But the onus is on the technology providers and the manufacturers to make sure the machine runs and it gets the job done.

Dan Miller can be reached at dan.miller@dtn.com.

Follow him on X @DMillerPF.

P[L2] D[728x90] M[320x50] OOP[F] ADUNIT[] T[]
P[R1] D[300x250] M[300x250] OOP[F] ADUNIT[] T[]
P[R2] D[300x250] M[320x50] OOP[F] ADUNIT[] T[]
DIM[1x3] LBL[article-box] SEL[] IDX[] TMPL[standalone] T[]
P[R3] D[300x250] M[0x0] OOP[F] ADUNIT[] T[]

Dan Miller