Skip to main content
Research

Robots on the Farm

Two new studies reveal the costs and benefits of automating farm production processes and enacting policies intended to benefit workers.

a woman milking a cow
Photo by Marc Hall

Heatwaves. Wildfires. Pests. Disease. Weeds. Floods. Droughts. None of these are what worry farmers most. What keeps farmers up at night is labor, according to a North Carolina agricultural stakeholder workshop held earlier this summer.

“You might think that most farmers see pests or weeds or weather events as their biggest concerns,” says Alejandro Gutierrez-Li, an assistant professor in the Department of Agricultural and Resource Economics at NC State University and an affiliated faculty member of the N.C. Plant Sciences Initiative. “But for the biggest issue to be one that could be resolved with either robots or immigration policy … that was kind of surprising.”

A word cloud representing the challenges identified by agricultural stakeholders in North Carolina. The larger the word, the more common the challenge.

Robots can tolerate adverse conditions, like heat and wildfire smoke, and can operate for longer hours. So what’s stopping farmers from using more automated technologies? Gutierrez-Li’s recent studies in Wisconsin, Minnesota and Washington reveal a more complicated truth about automating farm labor and the costs that come with it.

“Contrary to popular belief, mechanization or robotization is not necessarily a seamless linear process. There are pros and cons,” Gutierrez-Li says.

Dairy Debates: Robotic Milking in Wisconsin and Minnesota

Dairy has a labor problem. As a year-round industry, it does not qualify for H-2A visas, which are issued to temporary foreign workers allowed to work in the U.S. as seasonal farm laborers. But cows have to be milked every day, multiple times a day.

Faced with a lack of labor, the dairy industry in the Midwest has invested heavily in automated milking systems (AMS), which use robotic arms and sensors to attach cups to a cow’s udder for “hands-free” milking. Gutierrez-Li and his collaborators recently conducted a survey of dairy farmers in Wisconsin and Minnesota to determine their satisfaction with the technology.

According to the study, many of the surveyed farmers were satisfied with the milking technology, having been able to increase productivity, hire fewer people, plan more efficiently, make the cows more comfortable, and stress less about managing and retaining labor. 

“Before AMS, we were milking 180 cows […]. All labor included was 18 man-hours/day. That’s everything. Breeding, feeding, bedding, milking calves, heifers, everything! After AMS, it’s 12 man hours for the same work, plus 80 more cows and 10x more milk,” wrote one survey respondent.

“For our farm, it was either sell the cows or put in an automated milking system.”

Another respondent wrote: “For our farm, it was either sell the cows or put in an AMS.”

However, there are also clear cons to installing AMS. Those unsatisfied with AMS mention the difficulty and expense of maintenance and repair, as the equipment often breaks down without a trained person on hand to fix it.

One farmer responded: “We’ve had robots for nine years. Certainly, we cannot milk all the cows ourselves (just me and my husband), but the monthly bill is extreme. We can fix some things ourselves, but they have come [with] something wrong. We find them very frustrating.”

Though the stakeholder workshop revealed labor as the biggest concern for most farmers, the cost of keeping their farms running was a close second. And while automation may address labor concerns, it comes with its own heavy financial investment. Gutierrez-Li also notes that there are, of course, other considerations like commodity prices, animal welfare, sustainability, and farm succession plans that factor into the decision-making process.

Apple Harvesting: A Core Issue in Washington State

Apples are to Washington State what sweetpotatoes are to North Carolina. Another similarity: both states are in the top five for employing H-2A seasonal farmworkers. 

Gutierrez-Li recently conducted in-person surveys of apple harvesters in Washington for a forthcoming study that will be published in the Journal of the Agricultural and Applied Economics Association. He was instrumental in developing a survey in Spanish to better understand what the farmworkers valued most as beneficiaries of the H-2A visa program. Surveys of farmworkers in Spanish are uncommon, Gutierrez-Li says.

The survey found that farmworkers ranked the ability to take time off from work, to travel home to see family for instance, far higher than any other benefit. Somewhat surprisingly, the survey found that overtime pay was the least valuable benefit, which is a policy that the state of Washington recently implemented for H-2A workers.

The survey found that overtime pay was the least valuable benefit [to workers].

“The workers perceived that more regulations that are put in place to, in principle, financially benefit them will instead push farmers away from hiring more workers and incentivize them to invest more in mechanization and automation, putting workrs’ jobs at risk,” Gutierrez-Li says.

This study served not only to reveal farmworker preferences, but also to highlight the unintended consequences of well-meaning policies. These findings, while specific to apple farmers in Washington, can inform policy changes to better address the needs of all farmers and farmworkers.

A Shift in Labor

Making changes where labor, cost and automation intersect is a long game, Gutierrez-Li says. For technologies to become affordable enough to be adopted, they have to be created, refined and scaled, which takes decades. The first step is to develop a better understanding from farmers about which processes are mechanizable and whether farmers would find value in investing in automated technologies.

“We can’t just think about ag labor shortages as manual labor,” Gutierrez-Li says. “As we mechanize the production processes, which will take decades, other kinds of highly skilled workers are going to be needed.”

“As we mechanize the production processes, which will take decades, other kinds of highly skilled workers are going to be needed.”

New technologies relying on AI are generating large amounts of farm-level data. However, such information does not necessarily help farmers right away if they do not have the skilled workforce required to interpret it and suggest practical and implementable changes to increase efficiency. 

In the meantime, the immediate solution is one of policy reform. Gutierrez-Li says that immigrants will continue to be the largest source of agricultural labor for the foreseeable future, so optimizing policies like the H-2A program to meet the needs of farmers and farmworkers is crucial.