John Deere’s Robotic Tractor Is the Result of Years of Investment in AI-Powered Farming
When John Deere debuted its first autonomous tractor at CES 2022, it signaled that a new era of AI and robotic farming is soon upon us. While some companies have been talking about autonomous tractors for a long time, it is a completely different story when the largest US manufacturer of agricultural equipment shows that this is the future.
Still in the trial stage, early versions of the 8R are now being tested by what the company describes as its “paying partners.” But since it won’t be long before the final production model of the autonomous tractor rolls off the production line, I thought it would be a good time to sit down with one of the company’s computing directors, Chris Padwick, Director of Computer Vision and Machine Learning at Deere’s Blue River Technology division, to get an idea of how the company got to this point.
According to Padwick, since being acquired by John Deere in 2017, Blue River has helped accelerate the agricultural machinery giant into precision agriculture with its “see and spray” computer vision technology. The technology, which allows the farmer to apply a highly targeted herbicide to weeds in row crops, will first be rolled out on John Deere’s See and Spray Select in 2021.
The program makes it possible to do green-to-brown application, which is the application of a non-residual herbicide to plants at the “pre-emergence” stage. This application of precision herbicide application allows farmers to move from blanket application of herbicide to crops using older technologies such as plant mounds to precision application that can reduce the amount of herbicide used by 77% or more.
But it wasn’t until the latest incarnation of the technology, which uses green technology, that the benefits of Blue River’s investment in deep computer vision with learning capabilities (the company started researching it in 2016) were fully realized. With See and Spray Ultimate, farmers can spray herbicide in different crop seasons, powered by advanced neural network-powered computer vision that can distinguish between weeds of the same color as plants.
“If plants are touching each other, then all of your traditional computer vision methods for image processing – like morphology and erosion and template matching – kind of break down,” Padwick said. “It’s really impossible to build a system without these that can work with 95 percent accuracy or more.”
While Blue River helped John Deere accelerate its move to AI-powered harvesting, Padwick pointed out that the farm machinery company had already invested heavily in the technology even before Blue River arrived. By 2019, the company was processing 5 to 15 million measurements per second and had begun using computer vision to assess grain quality. According to Padwick, much of that work was based on work John Deere had done before Blue River arrived.
And today, all the combined skills and data gathered from across John Deere’s various AI efforts are helping the company create its first autonomous tractor, the 8R.
“In general, in all of our machine learning projects, we’ve tried to embrace the idea that all data is good data,” Padwick said. “We may have sprayer data from See and Spray collected from cameras that can help train an autonomous tractor. An autonomous tractor has different cameras, different geometry, and can collect data with different modes and different sensors, but that data collected from other projects may be useful for that training.”
When asked if he thinks farmers will embrace autonomous tractors, Padwick believes the answer is yes. He points to the rapid adoption of See and Spray as an example of how farmers will quickly adopt new technology that will help them do their work faster and at lower cost.
“I remember every time we did a See and Spray customer demo where we invited several farmers into the field, and they could watch the machine in action and give us feedback on the results. The overwhelming feeling was, ‘Wow, I knew this was coming, but I thought it was ten years away. You guys are showing me the future today.’ So I think the customer experience has been great and they are very hungry for innovation. “
Padwick says that once early adopters start using new equipment like the 8R automatic tractor, the word will spread to farmers, leading to more adoption.
“What I think is going to happen here is you’re going to see that other people are going to be very excited about the technology and adopt it, and the word will spread in the coffee shops,” Padwick said. “That’s how a lot of these products are sold, not through fancy marketing, presentations, or cool YouTube videos; it’s coffee shop conversations. And when people see that other farmers are starting to use a self-propelled tractor and they’re getting a good profit from it, that will create adoption. Because really, it’s a network of loyalty.”
