How John Deere developed its autonomous tractor

0


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


John Deere unveiled a £ 40,000 autonomous tractor at this year’s Consumer Electronics Show, which should be available in stores by the end of 2022.

The system uses six pairs of stereo cameras, combined with GPS guidance, to drive a Deere 8R tractor with a chisel plow and the ability to tow other equipment. A farmer can start the tractor with a swipe of a swipe on a smartphone app, and then go away to spend time with family or other business with warnings of anomalies the software cannot handle. While working, the tractor can also collect data on the condition of crops, the condition and moisture content of the soil, and other measurements.

Deere & Company’s chief technology officer, Jahmy J. Hindman, hailed the announcement as a milestone in agricultural productivity. “Until recently, agriculture was all about doing more with more – more horsepower, more input, more hectares – but the new digital age is changing all that. The last decade has been about doing more with less and farmers Providing additional tools to make better informed decisions, ”he said.

Above: Willy Pell of the John Deeres Blue River Technology Group with the self-propelled tractor presented at CES.

Image source: David Carr

Self-driving cars have been an integral part of the CES for several years. At this year’s fair, exhibitors offered components such as LiDAR sensors (Light Detection and Ranging) for autonomous driving and driver assistance systems. The Deere tractor doesn’t use LiDAR, however, and in general, the company’s technologists have been unable to port hardware or software from the self-driving car world to its application. As a computing environment, a tractor is fundamentally different from an automobile in terms of vibration, temperature, dust, and other challenges. For example, the stereo cameras are Deere’s own design, he said.

Hindman said that Deere can start with a few standard components, but all of them must be customized for its applications. For example, the tractor uses Nvidia Jetson Xavier GPUs, but with a custom assembly for passive cooling in a dusty environment where traditional computer fans would not be practical.

“As for the software side of things, it’s all new to us and everything,” he said.

To be able to explore uses of AI in agriculture, Deere spent $ 305 million in 2017 to acquire Blue River Technology.

Blue River provided the technology behind Deere’s see-and-spray herbicide application technology that can reduce the amount of chemicals sprayed on a field by approximately 80%, saving farmers money and promoting more sustainable agriculture. See and Spray is also based on computer vision with stereo cameras and distinguishes between plants and weeds when a sprayer passes over them to ensure chemicals are only sprayed on the weeds.

While Deere’s application is custom, it certainly benefits from more general advances in AI, such as shared architectures for deep neural networks, said Willy Pell, vice president of autonomy and new projects at Blue River. The Deere subsidiary has about 30 employees, he said.

For the autonomous tractor project, Deere technologists loaded prototype tractors with all sorts of sensors and ran them across fields, logging data to see which provided the most useful information, Pell said. “We ended up on stereo cameras, which wasn’t intuitive because we thought we were going to do LiDAR,” he said. Lidar proves to be great for detecting other vehicles on the road ahead of you, along with their distance, direction, and speed, but less relevant for a tractor navigating acres of open farmland.

In contrast, stereo cameras produce good depth perception, and Deere’s software is able to break the image down into multiple views – the raw image, a depth map, and a pixel-perfect classification of each part of the image, differentiating between the ground and the sky, and between one Crop and a foreign object that may have fallen on the field.

The tractor is programmed to deal with everyday occurrences, such as the need to drive around animals that might cross a field, but it also includes an anomaly detection system to help it cope with the unexpected. For example, if it detects an object on its way that is not included in its training data, it simply stops and signals the farmer to decide what to do.

Deere emphasized the care with which it was ensured that the machine can work autonomously but also safely.

This is an example of where a self-driving tractor has an advantage over a self-driving car. “If we come across an abnormal object, we stop – we don’t have to worry about another driver stopping us,” said Deanna Kovar, vice president of production and precision farming systems at Deere.

VentureBeat

VentureBeat’s mission is to be a digital marketplace for technical decision makers to gain knowledge about transformative technologies and transactions. Our website provides essential information on data technologies and strategies to help you run your organization. We invite you to become a member of our community to gain access:

  • current information on the topics of interest to you
  • our newsletters
  • closed thought leadership content and discounted access to our award-winning events such as Transform 2021: Learn more
  • Network functions and more

become a member


Share.

Comments are closed.