Robots on the Farm

Photo: Monarch Tractors
A Monarch tractor at Wente Vinvards

Smart tech is becoming more prevalent on farms, which innovators hope will spur new interest in farming.

Share this article!

Agriculture faces an uncertain future in Oregon and the country at large. Low wages, climate change hazards like smoke and extreme heat, and the relatively high toll farmwork takes on the human body, have led younger people away from farm work. 

According to a 2018 report by farm lending group AgAmerica Lending, the average age of farmers in the United States is 59 years old, and farmers under the age of 35 account for just 9% of the labor force. 

A 2020 report by supply chain software company GEP Worldwide found stricter immigration policies have made owning a farm less profitable. 

The GEP report recommended farms adopt agrotechnology — including robotics, artificial intelligence and predictive software — to cut costs and increase profits. Recent advances in robotics, along with Oregon universities’ agricultural programs and sustainability incentives have poised Oregon to become an epicenter of robotic farming. 

These advancements will not only make farming more efficient, but could also cause farming to be seen as a more appealing career path to students coming out of college. 

“It is important we make farm work exciting again. Some people look at farm work and see it as drudgery,” says Praveen Penmetsa, CEO of Monarch Tractors, a California-based company that makes fully electric driver-optional tractors. “It’s a lot more exciting to become a fleet operator and get to work with the latest technology.” 

After a successful deployment to Wente Vineyards in California, Monarch deployed its first driverless farming tractor in Oregon at Hopville Farms in Clatskanie in July. This was the first driverless tractor deployed funded by the USDA Conservation Innovation Grant. 

Penmetsa says his goal with Monarch Tractors is to eventually deploy them globally, both to reduce greenhouse gas emissions and make sure everyone has access to safe, effective farming tools. But before that, the technology needs to be tested and perfected in real-world conditions. For that, the company will partner with Oregon State University to use the institution’s data gathering software to improve the tractors’ ability to collect and react to field data to improve agronomy.

Monarch Tractors are not the only way OSU is advancing agrotechnology. 

This year, 13 researchers from OSU’s College of Engineering — in partnership with the University of Washington — are participating in AgAID, a $20 million federal effort to develop artificial intelligence to address farming’s biggest problems: diminishing water, labor shortages, and extreme weather variations caused by climate change. 

The plan is to incorporate predictive technology into farming tools to be able to deliver outcomes specifically-tailored to individual fields. A feature that will become more necessary as climate change makes weather patterns more destructive and unpredictable. 

“These AI in the systems still needs a lot of improvement. It’s very much an art to get ripe grapes, for example, thirsty enough and feed them just enough to water the outcomes you need. Every field is different,” says Alan Fern, professor of computer science and lead investigator representing Oregon State. “One of the things working on is giving growers an interface to forecast the right line of irrigation based on their growing goals based on their field that becomes highly customized. A lot of it is setting up good data pipelines for what is there and what should be there.”

While seeding and tilling are able to be accomplished by AI units, things like pruning trees and picking apples are more difficult to program, since fine motor skills are required. 

For now, agrotech needs to learn from human experts in order to teach machines best practices. With aging experts retiring from the profession, passing on knowledge to artificial intelligence systems, rather than a new generation of farmhands, ensures the machines begin their deployments at a high level of functionality. 

“Part of our team focuses on human computer interaction. if you’re not designing with end users in mind it’s not going to get used. it has to be a user-first design process,” says Fern. “AI is not this magic thing that can solve all problems. it’s not going to replace all workers, it’s hopefully going to make workers lives better and more efficient.”

 To subscribe to Oregon Business, click here.