Soybean farmers use AI to raise their crop. I know AI, that’s artificial insemination that groups calving times.
My brain must shift gears. At the Soybean Symposium I learned AI also stands for artificial intelligence. Soybean growers told how computers drive planters, combines and much more.
I learned a new term: machine learning. I’d not heard that but like it better than AI for what soybean farmers do. This AI, or Big Data, allows farm equipment to apply variable farming methods. That’s needed as it’s a rare field that’s uniform across the acreage. Mother Nature isn’t consistent.
A farmer member on the producer panel told of his river-bottom fields. They vary from sand to loam to gumbo, one end to the other. Last fall yields varied from zero to 80 bushels in one pass.
Over time, farmers learn to farm variable fields. The low fertility parts don’t need as much seed. Variable fields also need variable fertilizer, herbicides and fungicides. Actually it’s variable everything.
A speaker from Bayer Corp. told how she writes computer programs for machinery learning. AI is here and useful.
A speaker from Purdue University told of her research using aerial drones to map fields and plant growth. It’s gotten down to knowing the field differences from one square meter to another square meter. This extreme micromanagement uses thousands of megabytes of data. No human mind can think and retain that data. Computers come installed on all kinds of farm equipment.
This type of farming has grown. Another farmer panelist told he’d just bought his third planter. He also said he liked to retain some control. “So I can stay awake, since I no longer have my hands on the steering wheel, I need something to do.”
Tractors have been driven by satellite guidance for some time. The machinery learning is taking over more tasks. Varying inputs across the field can cut the use of chemicals and reduce costs.
But, I didn’t hear any economics cited. Do increased yields offset the higher cost of machines learners?
The Symposium told the astounding potential of what can be done now, with equipment available. Of course, new stuff will come at a pace faster than I can keep up.
One of the first questions for the machine-learner guru: “Who owns the data?” Is it corporate collectors or the farmers?
Bill Wiebold, MU Extension agronomist also known as our “Soy Doc,” was early user of drones to scout soybean test plots at MU Bradford Farm. He also monitors variety tests across the state. In cooperation with local farmers he has corn, soybean and wheat plots. Keeping track of all those is a huge effort.
He’s upgrading drones to do more research chores. Plus drones can be adapted for precise applications of inputs, weed killers to fungicides. There’s variability across his small plots.
Shifting gears only slightly, last weekend, I heard a complex dialog between two MU professors, one a biochemist and the other an economist. This is a regular happening, every Saturday, when professors preview what they think and do.
This week, they covered a new way of using statistics. It’s called “likelihood statistics.” Not far into debate, my mind trained on old statistics decades ago was saying, “No, this can’t be. No way!” But, it may be that my brain needs a refresh.
Statistics, like new varieties, are upgraded. All of this requires new ways of thinking. Famers will be challenged to keep up, as never before. Actually in daily news about political polling, I may already be hearing “likelihood” analysis of how voters will vote.
Will U.S. voters become as smart as soybean farmers using AI? That’s artificial intelligence, not insemination.
At the Symposium, I pondered how to explain new AI to soybean farmers. Maybe their kids know computers and statistics.
An aside: Lunch table talk was more about growing hemp than beans. Will hemp become our No. 1 crop?
