Field of AI: Startup Helps Farmers Reduce Chemicals and Costs

Israel’s Greeneye is working with Midwest U.S. corn and soybean farmers.
by Scott Martin

Brad Janzen, a farmer based in Henderson, Neb., with more than 4,000 acres of corn and soybeans, knew he had to do something about herbicide-resistant weeds.

Farms everywhere faced the rising costs of switching to new herbicides and the increased soil contamination.

“Invasive weed species have inherited the ability to resist herbicides used in row-crop agriculture, so we knew a new approach was required if we were to continue to use these herbicides in our fields,” he said.

Amid this environment, Nadav Bocher, Alon Klein Orbach and Itzhak Khait were developing AI for precision spraying of crops for U.S. markets to help reduce the application of herbicides. They founded Greeneye, based in Tel Aviv, in 2017 and now offer AI-driven smart sprayers for farms.

Greeneye has landed $12 million in funding to date from JVP, Syngenta Group and other investors, including Eyal Waldman, the former CEO of Mellanox Technologies, acquired by NVIDIA in 2020.

Driven by the NVIDIA Jetson edge AI platform, Greeneye’s system can detect and precisely squirt weeds in milliseconds, reducing the use of herbicides by as much as 90 percent and helping reduce water and soil contamination along with costs.

“We also collect an enormous amount of high-resolution data from the field to provide insights to farmers and other stakeholders,” said Bocher, the company’s CEO.

The startup is now working with corn and soybean farmers in the midwest U.S., aiming for wider commercial launch of its services next year.

Greeneye is a member of NVIDIA Inception, an accelerator program that helps startups in AI and data science with tools, training and technology expertise to get to market faster.

Reducing Chemicals Usage

Farmers typically either spray more or switch to other, often more expensive, herbicides. And the industry practice for decades has been to blanket spray crops. None of this is good for the land, and herbicide costs for farmers have climbed, said Bocher.

To address this problem, Greeneye and others are developing AI for precision spraying of herbicides. This new crop of agtech upstarts uses AI to benefit farmers and the environment, and includes companies such as Bilberry, FarmWise, SeeTree, Smart Ag and John Deere-owned Blue River.

Greeneye’s proprietary dual-sprayer technology enables farmers to precisely target herbicides on individual weeds as well as apply residual herbicides more broadly as needed.

“Greeneye’s selective spraying system allows us to deliver more herbicide to just the weed, while lowering total herbicides used,” said Janzen.

GPU-Driven Farming

Tractors outfitted with Greeneye’s AI can detect weeds for treatment at a rate of 40 frames per second while traveling at 13 miles per hour, thanks to the NVIDIA Jetson AI platform.

This type of millisecond-speed inference for weed recognition and treatment was achieved by tapping into TensorRT optimization from the NVIDIA JetPack SDK.

“Weed infestation in a field can be as little as 10 percent, yet we go and spray the entire field,” said Bocher. “So, fundamentally that’s what we’re looking to change as a company.”