It’s harvest time. Sure is different. The rechargeable self-driving harvesters started out this morning. Looks like the crop will come in just as predicted by the AI. The satellite images confirm it but the machine vision inspection also verified each plant during the season. The future of farming is already here…
My wife volunteers for a nature conservancy in one of our big municipal parks. One of her projects is planning a Fall Festival for families with hayrides, pumpkins – the whole shebang. Most of us live in urban enclaves but we still have romantic images of life on the farm that seems to be frozen in the mid 20th century. Picture that bucolic family farm worked by Jonathan and Martha Kent (Kal El’s foster parents – wink, wink).
Welcome to the 21st Century where reality is more nuanced. First, surprise! – Family farms accounted for nearly 99 percent of U.S. farms – Hard to believe? I heard evil corporations were taking over the American farmland. Not so, small family farms account for 90 percent of all U.S. farms. Here’s the nuanced part. Large-scale family farms account for about 3 percent of farms but 42 percent of the value of production. Finally, the 1% or so of non-family farms account for almost another 11% of production.
What’s the take away? One, there are still lots of family farms – that’s good. Two, most of the food you eat comes from big farms, regardless of ownership. Farming has always succumbed to the economies of scale. Mechanization revolutionized farming and created a virtuous circle where expensive capital equipment required more land and production to justify it and the bigger farm justified that more capital equipment could be deployed.
That cycle does reach a point of diminishing returns. Today, we are in the next phase where we couple the power of digital technologies to farming to enhance productivity, reduce herbicide and pesticide use and minimize risk. Farmers face a challenge requiring this. Today an American farmer feeds about 155 people. Depending on what global population growth you pick soon he/she will need to feed 264 people.
What kind of tech are we talking about? Let’s start with an overall bird’s eye, or maybe god’s eye, view. From 250 miles up, a start-up called Planet orbits a “flock” of 149 satellites to provide daily observations of every spot on the earth. Ever hear of it? Why not, that seems like a lot of hardware. They must have had a ton of rocket launches. Actually they didn’t. Leveraging digital technologies the company has pioneered what it calls CubeSats which are roughly 4in x 4in x 12in in size and weigh only 8.8 pounds.
1000 times less in weight than predecessor earth observation satellites these miniature wonders are a testament to the phenomenal advances in electronics that we all experience in our smartphones. At that weight scores can piggyback on launches of bigger spacecraft. From these little guys pour a wealth of information used by companies like Vinsight and FarmShots to offer detailed farm field conditions and predicted yields.
Now, let’s zoom down onto the field. The satellite findings can be fed into self-steering tractors with GPS to apply more or less fertilizer as needed to optimize yields while minimizing fertilizer consumption. Let’s get even closer and look at each plant. John Deere just bought the machine vision company Blue River. Each plant is individually examined to determine if it is desirable. The unwanted are directly squirted with a tiny amount of herbicide to remove them. This sort of precision agriculture results in up to a 90% reduction in herbicide.
We are not done yet. A growing season can be a tough time when plants are subject to all kinds of hazards. Monsanto has dropped over a $1 Billion in investment in a network of sensors and applications for the farm called Climate Field View, which covers 92 million acres, today. They are attempting to build an ecosystem of developers who will utilize the platform to create all sorts of new applications for farmers. (It’s like an Amazon Cloud for agriculture.)
The ultimate goal is to teach AI’s to farm. Monsanto and BASF are applying machine learning and computer vision to the traditional challenges of farming. BASF has already put into trial a program called Maglis. Think of it as an increasingly smarter decision support tool for the farmer.
The family farm is still here but only 2% of the U.S. population are farmers. It’s a tough business: too much rain or not enough rain; pests, weeds, and labor shortages. It’s amazing that anybody wants to do it. But now, tiny spy satellites, robot tractors, AI and machine vision can all give them a leg up. Information technology and digital tools are changing how we do one of mankind’s oldest activities, farming.
Image credit:The Case IH Autonomous Concept Vehicle (Case IH / Handout)