How John Deere created its autonomous tractor

Hear from the CIO, CTO and other C-level and senior executives on data and AI strategies at the Future of Work Summit on January 12, 2022. Learn more


Adding a new dimension to the self-driving vehicle industry, John Deere unveiled a 40,000-pound autonomous tractor at this year’s Consumer Electronics Show that it says will be commercially available by the end of 2022.

The system uses six pairs of stereo cameras in conjunction with GPS guidance with the ability to operate the Deere 8R tractor with a chisel and to pull other equipment. A farmer can put a tractor to work with a swipe of a smartphone app and then walk away to spend time with family or attend another business, use the app to plow a field or do other work to monitor the progress of the tractor – and Warnings of discrepancies to receive do not know how to handle the software. While it is working, the tractor can also collect data on the farm about crop health, soil health and moisture and other parameters.

Jahmi J., Chief Technology Officer, Deere & Co. Hindman celebrated as a landmark declaration for farm productivity. “Until recently, agriculture has always had more to do with more – more horsepower, more inputs, more acres – but the new digital age is changing all that. Is about, ”he said.

A picture of a man leaning against an autonomous tractor by John Deere.

Above: Willie Pell of John Derena’s Blue River Technology Group with a self-driving tractor at CES.

Image Credit: David Carr

Self-driving cars have been a major part of CES for many years, and this year’s show featured exhibitors offering components such as LiDAR (light detection and ranging) sensors for autonomous driving and driver assistance systems. However, Deer Tractor does not use LiDAR, and in general, the company’s technologists were not able to port hardware or software from the world of self-driving cars into their applications. As a computing environment, tractors differ fundamentally from automobiles in terms of vibration, temperature, dust and other challenges. So stereo cameras, for example, have their own design, he said.

Hindman said Deere is able to get started with some off-the-shelf components, but it must always be customized for its applications. For example, the Tractor Nvidia Jetson Xavier uses a GPU but with a custom assembly for passive cooling in dusty environments where traditional computer fans are not practical.

“When it comes to the software side of things, we’re all new and new,” he said.

To put itself in a position to find AI applications in agriculture, Deere spent $ 305 million acquiring Blue River technology in 2017.

Blue River provided the technology behind Deere’s C & spray technology for the use of herbicides, which it says can reduce the amount of chemicals sprayed on farms by about 80%, saving farmers money and promoting more sustainable farming. Based on computer vision with stereo cameras, look and spray distinguish between plants and weeds as chemicals are sprayed only on the weeds to ensure the sprayer passes over them.

While Deere’s application is custom, it certainly takes advantage of more general advances in state of the art for AI, such as general architecture for deep neural networks, autonomy with the Blue River, and new ventures, said Willie Pel, vice president of new ventures. Deere’s subsidiary has about 30 people, he said.

For the autonomous tractor project, Derena technologists loaded prototype tractors with all sorts of sensors and drove them into the field to see if the most useful information was provided, data logging is done, Pel said. “We landed on stereo cameras, which was vague because we thought we were going to do LiDAR,” he said. Lidder proves to be excellent for detecting other vehicles on the road ahead with their distance, direction and speed, but is less relevant for tractors navigating acres of open farmland.

In contrast, stereo vision cameras generate good depth perceptions, and Deere’s software is capable of splitting an image into multiple views – a raw image, a depth map, and a counter-pixel classification of each part of the image showing the difference between ground and sky. Between a crop and a foreign object that may have fallen into the field.

The tractor is programmed on how to handle common events, such as the need to navigate around animals crossing the farm, but it also includes an anomaly detection system to help cope with unforeseen events. For example, if he finds something in his path that is not in his training data, he just stops and gives a hint to the farmer to decide what to do.

Deere stressed the care they took to ensure the machine could operate safely, even autonomously.

This is an example where a self-driving tractor has an advantage over a self-driving car. “When we find something inconsistent, we get stuck – we don’t have to worry about the rear end by another driver,” said Dana Cover, vice president of production and precision agricultural systems at Deere.

Venturebeat

VentureBeat’s mission is to become a digital town square for technical decision makers to gain knowledge about transformative technology and practices. Our site delivers essential information on data technologies and strategies so you can lead your organizations. We invite you to access, to become a member of our community:

  • Up-to-date information on topics of interest to you
  • Our newsletters
  • Gated idea-leader content and discounted access to our precious events, such as Transform 2021: Learn more
  • Networking features and more

Become a member

Similar Posts

Leave a Reply

Your email address will not be published. Required fields are marked *