Down to the Individual Grain: How John Deere Uses Industrial AI to Increase Crop Yields Through Precision Agriculture

Recently, The Verge spoke with Jahmy Hindman, CTO at John Deere, about the transformation of the company’s farm equipment over the last three decades from purely mechanical to, as Jahmy calls them, “mobile sensor suites that have computational capability.” This is in service to John Deere’s “smart industrial” strategy. More than just selling a piece of equipment, smart industrial is about providing the whole system (equipment, data, analysis and automation) that farmers need in order to provide individualized care (exact amount of water, nutrients and pesticides) at scale to each of the tens of thousands of plants per acre (multiplied by thousands of acres per farm), leading to greater yields and lower waste.

At this year’s Data + AI Summit (DAIS), Gregory Finch (Senior Principal Software Engineer, Intelligent Solutions Group) and Jake Sankey (Technical Product Manager, Enterprise Data & Analytics Platforms) from John Deere went in-depth about the data platform that makes this possible during their manufacturing keynote. As the amount of data generated by equipment doubles or triples per year, Deere needed a data platform that could handle this scale of data now and in the future, easily integrate new data sources (e.g., weather) and then unify it so that different downstream teams -- like sales, service, or engineering -- could improve customer results.

As Jake explained, “our technology stack is really vast...It consists of onboard and offboard components. On the onboard side, we have sensors, tons of them. We have vision systems, guidance systems and wireless connectivity. Offboard we have cloud infrastructure and storage and scalable services that allow us to receive and process and analyze all that data. This stack is what enables us to help our customers be more productive and more successful.”

As an example, he points to the X9 Combine (the machine that harvest grain crops) where, “cameras continually monitor images of the grains down to individual kernels as they're taken up the combine's elevator and dumped into the tank. We use machine learning to analyze grain quality and automatically adjust the operating parameters of the machine if any damage is detected to the grains.”

These sorts of advancements don’t just help the farmer, but have broader social benefits as well. Through precision agriculture, farmers can reduce chemical use by 70%, reducing environmental impacts of pesticide overuse.

Throughout this keynote, Jake and Greg talk about how a 184-year-old enterprise is leading the transformation of the industry as data and artificial intelligence (AI) become more prominent tools of the trade—from execution on the shop floor to how things work in the customer’s hand.