Michigan farmers produce roughly 300 different commodities, cementing Michigan as, “the most diverse agricultural state in the country with a reliable source of water,” an expression attributed to Peter Lemmer, executive vice president and chief legal counsel for GreenStone Farm Credit Services. Among these commodities are specialty crops, including many fruits, vegetables, tree nuts, dried fruits and nursery plants.
Current challenges, however, are threatening various specialty crop growers, forcing some to decide whether they should switch which crops to produce or — in some cases — leave their industries completely. One of the issues is labor, specifically the rising cost of foreign labor and lack of domestic labor within U.S. agriculture.
To safeguard Michigan’s diverse agricultural landscape, a Michigan State University agricultural
engineer is advancing technologies that enhance mechanization and automation during harvest and production, providing growers with solutions to withstand ongoing labor trends.
Yuzhen Lu, an assistant professor in the MSU Department of Biosystems and Agricultural Engineering, received a $197,000 grant in 2024 from the USDA National Institute of Food and Agriculture’s (USDA NIFA) Hatch Multistate Research Fund, administered through MSU AgBioResearch, to develop a vision-guided, selective harvesting mechanism for green asparagus.
Michigan leads the U.S. in asparagus production, harvesting roughly 20 million pounds annually, according to the Michigan Ag Council.
The process to harvest the stalky vegetable, though, is tedious.
Asparagus spears grow upward out of the ground with their crowns pointed to the sky. Workers sit on tractors called asparagus carts that are driven over rows of asparagus. Sitting low to the ground on these carts, workers “snap” the spears from the ground at the length at which they’re desired — a process unique to Michigan. With the ability to grow an inch per hour on hot days, asparagus needs to be harvested daily, sometimes multiple times per day.
In talking with MSU Extension educators and specialists, Lu found that labor costs alone can equal 55% of the total cost for asparagus production.
Jamie Clover Adams, executive director of the Michigan Asparagus Advisory Board, agreed that 55% is an accurate average representing labor costs within total production cost, adding that the number can be higher for some farms.
“Asparagus is a very labor-intensive commodity, and for some farms, the cost of labor is more than 55% of the total production cost, but that’s a good average,” Clover Adams said. “Because that price is only going to go up, it makes mechanization important for the long-term future of the industry in Michigan, and quite frankly, the U.S.
“Michigan is the No. 1 producer of asparagus — not because the state has grown in its production, but because every other state has declined, and that’s due to labor and cheap imports. So, the labor-saving device Dr. Lu is working on would really help us in the long run.”
Lu said the first attempts to mechanize the harvesting of asparagus date back to the early 1900s. Research in this area boomed during the mid-1900s as several automatic harvesting prototypes were developed, but ultimately none were effective enough to move forward.
After meeting with growers and industry professionals, Lu said he learned how advancing this technology would benefit the industry in its current state.
“As an ag engineer, I felt compelled to do something for the asparagus industry, especially during the early stages of the proposal process when I learned about the urgent need for automation in this space,” Lu said.
The selective harvesting technology Lu and his team, including MSU Horticulture associate professor Zachary Hayden, are developing would allow mature asparagus to be harvested without causing damage to immature asparagus not ready to be picked.
Powered by a 3D camera, a vision-guided system would inform the automated machine as it’s driving over rows of asparagus on which ones to pick based on maturity status and desired traits, such as spear length and shape.
“The goal we eventually want for this technology is to have it accomplish what human pickers can do because harvesting asparagus is selective and requires decision-making,” Lu said. “Asparagus grows at different rates within the field, so if harvesting technology doesn’t have the ability to be selective, there’ll be a risk that 50% or more of a yield may be lost.”
Lu and his team are currently building out the database — or as Clover Adams described it, “the brain” — for the vision-guided system to operate and detect different topographic features of asparagus.
In addition to selectivity, another quality Lu’s team will work to impart within this technology is efficiency. Lu said the design of the harvesting component needs to be simple, noting that he and his team might take ideas from the mid-1900s and adapt them for today to mimic human efficiency.
Clover Adams said she agrees this technology needs to be efficient to be practical.
“I look at Dr. Lu’s work as a way to catapult us forward faster than we otherwise would have because other places developing machines are developing them for where they are and what they need, and they’re just not fast enough for us,” she said.
The project is currently funded through 2026.
In a separate proposal supported by the USDA NIFA Hatch Multistate Research Fund, Lu received $200,000 in 2023 to aid another specialty crop important to Michigan: blueberries. Funded through 2025, Lu’s team is creating a mobile machine vision system for in-orchard blueberry management.
Using a phone app, growers would be able to detect the maturity and count of blueberries by scanning their orchards with the cameras on their phones. Lu said this could become a beneficial tool to validate when growers should harvest.
“In practice, growers rely on blue fruit percentage to schedule their harvesting,” he said. “Without technology, growers are coming through and checking the color of the blueberries manually. Doing this is time costly and inefficient. My team thinks if growers can take pictures of their canopies and let the camera systems tell them how many berries have reached maturity level, it would be much more efficient.”
Lu and his team, including MSU Horticulture assistant professor Josh Vander Weide, have already published a paper with preliminary results regarding the app, but they’ll spend the next year testing it with hopes of releasing it soon to blueberry growers.
This story was originally published on AgBioResearch.
