Posted on

slanted farms seed company

Slanted farms seed company
Jones is CEO of bext360. His company’s robot could help fix this problem, but we’ll get to that later. First, let’s get back to the ground.

It starts with a seed: A Food Robotics Explainer for Kids

It starts with a seed. An apple seed, a tomato seed, some type of seed gets planted in the ground. Then that seed grows. And it grows. Slowly, the plant pierces through the dirt and into the light. Weeks to months later, it bears fruit. What was once an apple seed is now a tree standing tall with countless of apples on it. That tomato seed has become a plant, waist-high, bearing dozens of ripe tomatoes. The fruit gets picked and, when the harvest is plenty, ships out to warehouses where a restaurant or grocery buys it. A cook takes the fruit, cuts it up, then puts it in a salad for you and a salad for me.

But the process is changing. Dan Steere is CEO at Abundant Robotics, a company that makes robots. He says, “Everything about farming, everything you can look at about farming, there’s major technologies coming in the next 10 years to make each part of farming more efficient, more productive, and hopefully healthier and less expensive.” By time you’re an adult, the entire food cycle will likely be robotic.

That’s almost the case now. If we go back to our seed, we’ll see robots are already involved in the very beginning: They show farmers where to plant.

Land on a farm is not level: It can have rises or holes or ditches in it. Plowing can even the ground out some, but never completely. If there’s a creek running through the middle of the field, there’s always going to be land around that creek where it’s difficult — or impossible — to plant. You want a creek because that means access to water. It also can mean more fertile soil. But a field’s uneven nature can make it harder for a farmer to maximize his land. With math, he can calculate how many seeds should be planted where to produce the largest harvest. But land also changes, so the work has to be redone and redone and redone.

To make things harder, every field has some parts that naturally are less fertile than others. Soil quality impacts the volume of food produced. It also affects how good that food is. Or as Daniel Jones puts it, “If Terena had a farm over here and Dan had a farm here, basically they should be producing the same amount of [crop] per hectare…but [Terena’s] quality was much better, and therefore [she] got a lot more money for [her] crop.”

Jones is CEO of bext360. His company’s robot could help fix this problem, but we’ll get to that later. First, let’s get back to the ground.

Theo Pistorius founded DroneClouds in South Africa. They’re one of many companies that use drones to help farmers know where to plant. Drone is slang for unmanned aerial aircraft, or UAA for short — a flying robot — and the drones that DroneClouds uses have cameras on them. Five cameras, in fact. Pistorius says, each one “is essentially a camera on an iPhone, but just a very specialized, aerial iPhone, with a very specialized, calibrated camera.”

First these cameras take pictures of the land from overhead. The pictures show area and vector, but they also reveal soil variation and irrigation problems that the farmer may not know about. They can even show where bugs or fungus might be! All of this information is tremendously helpful when deciding where to plant. Farmers have used aerial photos to help with this since the 1980’s, but because you don’t have to pay a pilot or maintain a plane, it costs less for drones to take the same shots. They can take them more often, too.

Once the farmer’s figured out where to plant, using a UAA from DroneClouds or a competitor, it’s time to get that seed in the ground.

This is where a self-driving tractor comes in. Depending on where you live, you may have seen a self-driving car. Some states — like California — allow them, but others don’t. While governments decide whether to allow self-driving cars on the road, John Deere has been selling self-driving tractors for 17 years! Since tractors haul planters and setters — farm implements that put seeds in the ground — you can see that robots have been at work in the planting part of the process for a very long time.

So when will we learn about what’s new? Patience, young farmer. Like our fall crop, it’s coming.

Robots have helped the farmer figure out where to plant, and they’ve gotten the seed in the ground. But now that baby seedling is starting to push beyond the surface. That little spot of green means spring, and it’s time for the plant to grow.

Remember DroneClouds? Well, that drone is still in the sky, taking pictures. With those pictures, Pistorius says, “We do a bit of processing, which gives us various maps about the field and the [growing apple] trees. We then do analyses to interpret it for the farmer…How many trees are coming up, and where do the weeds cause the apple trees to struggle?”

To pinpoint problems, these pictures are analyzed against others the drone took of that same crop. This is called comparative analysis. Pistorius says it’s like running a race then comparing your time to earlier in the season. But runners also compare their time against other runners; that’s how school records are won. Farmers also compare pictures of their field to other farmers. This is signature-based analysis. As Pistorius explains, “The ideal pictures come from labs all across the world…[S]cientists look at thousands of these images before making an ‘ideal picture’…Every 4 years, scientists from the Agricultural Research Commission [in South Africa] meet with labs [in the US] (such as the American Agricultural Society), and take a bunch of signatures.”

Then day after day, our little plant grows. The sun rises and falls, and sometimes it shines and sometimes there’s rain. Finally, harvest comes — and with it, new, cutting-edge work in agricultural robotics.

For three years, Abundant Robotics has been developing a robot that picks apples. Three years? But picking apples is easy! Well, not if you’re a robot.

To understand why this is hard for a machine, let’s break down the apple-picking process: When you see an apple hanging on a tree, your eyes send a signal to your brain. Your brain processes the data in this signal — like the apple’s color or where it is on the tree — and, instinctively, you know the apple’s ready to pick. Your brain then tells your arm to reach out and your hand to pick it. You hold the apple like you would a bird — gently enough not to bruise it, but firmly enough that it doesn’t fall away.

So the first problem Abundant Robotics had to solve was inputting the right signals. “If you don’t have a good pair of eyes, it’s hard to do a lot of tasks in the real world,” Steere says, so the first thing his company had to do was give its robot “a better pair of eyes.” These eyes — and the way they connect to the robot’s brain — are called computer vision. Computer vision helps the robot see “images of every surface of an apple” says Steere, including “the size, the color quality, the weight, and any defects.”

Even with great eyes, the robot still had to learn how to drive through the field and how to pick apples without damaging anything. “[H]eavy animation damages the fruit. If you bruise the apple or cut the skin — any type of rough handling ruins the fruit,” Steere says. “Also as you pick it, it damages the trees.”

Then the robot must use vision and motor skills together. Think back to our apple-picking process: You have to know which apple to pick. You have to pick it quickly and gently. But what else? You have to be careful not to disturb other apples that aren’t ready. Steere says, “The vision has to both recognize fruit and it has to recognize whether it’s ripe or not in a fraction of a second.”

Now you see how this work could take years! “People have wanted to automate this type of agriculture for decades. It’s just never been possible,” Steere says. He also gives credit to the “thousands of engineers over a couple centuries working to solve these various problems”: “You can learn a lot by just sort of paying attention to what people have had to do in the past.”

And the work’s not done yet: Abundant’s robot still isn’t done. Great tech development is like farming — it takes patience.

If you’ve ever shulled peas, you know patience is also important in the next stage of the food cycle: Sorting. Once the fruit of a plant’s been picked, good harvest has to be sorted from the bad. That’s what bext360 does.

Instead of apples, their robot works with coffee cherries (coffee cherries hold coffee beans), cocoa, nuts, and cardamom. Jones says, “The farmer would harvest their coffee and place it in our machine and our machine would analyse it using machine vision…then the machine drops it through a visioning system where you can picture like a waterfall of cherries falling…and our system takes pictures.” It then uses these pictures to sort the good coffee cherries from the bad.

Machine vision is another phrase for computer vision. Even though Abundant and best360’s robots do different things, the same technology gets the job done.

Another thing both robots have in common is that it took more than computer vision for them to work. After vision to tell how to sort, the robot actually has to sort! Jones says, “Each collection process is about 30 kilograms. In scientific terms, it’s pretty impressive because 30 kilograms of coffee cherries is about 18,000 beans to give you an idea. So it does that in about 3 minutes. It sorts through, and it takes pictures of each one and analyzes every one…We’ll know everything about them in that split second that they fall through, technically in about 22 milliseconds…As they’re going through we have the ability to then do typical machine sorting…puffs of air will hit the beans and we can sort them into different bins.”

(Farmers put cherries from one section of the field in at a time. The robot catalogs this data with other information. which group that cherry came from in and shares its analysis with the farmer.) As the coffee cherry falls, the robot catalogs which group that cherry came from in and shares its analysis with the farmer. Earlier in this article, we promised to tell you how bext360 helps farmers figure out where to plant. That’s what farmers do with the robot’s quality analysis. Jones says computer vision collects information about “the quality down to every cherry for everything. The main things are size and color and density but we also can inspect for what they call intrinsic defects and extrinsic defects and disease, right? So we can give feedback to the farmers and correlate that together with their farming practices,” Jones says.

Picked, analyzed, and sorted, off our harvest goes to the warehouse — one day maybe in a self-driving semi. Maybe one day a self-driving tow motor will move the boxes off that truck and onto another semi heading to the restaurant or store. Amazon already has a grocery store just for employees that doesn’t have any stockers or check-out clerks: They’re all robots.

So now that our food has gotten where it’s supposed to, it’s time to meet Sally, the last robot in our story. She makes salads. And who made her? A company named Chowbotics, created by Deepak Sekarby.

Sekarby says, “There are cylinders inside the robot that are filled with prepped ingredients.” Chowbotics is pretty secretive about how Sally slices and dices the vegetables you put in her, but says you turn Sally on by pressing a touch-screen interface. This same interface also “customize[s] your salad by

And no, in case you’re wondering, Sally does not look like Rosie, the robot maid on “The Jetsons.” “Sally is a box with the robotic components on the inside,” Sekarby says.

Want your own Sally? You’ll have to save you allowance. One robot costs $30,000.

So there you have it — from start to finish, seed to salad, this is how robots are involved in the growing and preparing of food. You can already tell from reading this that there are still problems in this process that need kids like you to figure them out.

Steere says, “If there was ever a time [to be] a young person going into farming — this has gotta be one of the most amazing times in history. The kind of things that automation can do is going to continue to change and to evolve quickly, so more automation in pretty much any type of farming.”

So now that you have the patience to watch things grow and the knowledge it takes, what farming problems will you solve?

(A version of this article was initially published by Science News for Students.)