A commercially viable robotic strawberry picker cannot be available quickly enough for most growers today.

“We’ve left fields that were beautiful that we couldn’t get to,” said Andy McDonald, farm manager for strawberry grower Sweet Life Farms in Plant City, Florida. “It happens quite frequently these days, and it’s very heartbreaking. You just have to walk away.”

McDonald, who has left behind as much as 25 percent of his strawberry yield because of labor shortages, is one of several strawberry growers invested in a robotics effort that garnered national acclaim with a grant from the National Science Foundation (NSF) in December.

More than 20 percent of U.S. strawberry growers have a stake in Harvest CROO Robotics LLC, a company started by Florida berry grower and innovator Gary Wishnatzki and Chief Technical Officer Bob Pitzer in 2012. Investors have given $1.8 million to Harvest CROO so far, and the company is raising $3 million for its next version. The grant from the NSF could total as much as $1 million. This year, Harvest CROO added two full-time engineers to the team with GPS navigation expertise.

The company’s latest prototype harvester was tested in Florida strawberry fields this winter with hopes that its 16 picking heads would more than rival a crew of 30 human pickers.

Harvesting robotics is not a new concept to strawberries, although Harvest CROO’s planned approach may be. Wishnatzki, a third-generation grower and owner of Wish Farms in Plant City, Florida, said the harvesters will maneuver autonomously in existing in-soil fields and they will be leased to growers, alleviating the need for large capital investment and costly maintenance.

“Leasing would be better for all of us involved on the grower side,” said McDonald, who devotes about 400 acres to winter strawberries and also grows watermelon, cantaloupe and vegetables. “One of those machines will be a lot of money to buy. If you think about the mechanical upkeep, leasing through Harvest CROO and having their own specialized technicians to fix things and do the maintenance is going to be much more beneficial to us.”

Harvest CROO Robotics’ prototype incorporates a picking wheel and vision technology to identify and pick ripe strawberries, and company executives say they are holding the vision technology as a trade secret. The company’s goal is to lease strawberry pickers to growers for existing strawberry installations.

Getting noticed

Harvest CROO’s prototype was poised to halve the previous season’s picking time from 8 to 4 seconds per strawberry plant. The company’s goal is to use what it is learning to build an alpha version for 2018, moving closer to commercial introduction of the robotic service in 2020.

Globally, several programs tout their ongoing development of robotics to address different aspects of the labor-intensive crop, which requires harvesting every three days during its season. In some cases, developing harvesters are designed to run in soil-less environments or substrate farming systems.

“We recognize that industry leaders, including ourselves, are going to have to push forward technology projects,” said Hillary Thomas, director of Naturipe Berry Growers’ newer research and development department, in Salinas, California. Naturipe joined the list of Harvest CROO’s investors last year.

“Harvest CROO isn’t the only project we are using capital to drive forward in harvest automation,” said Thomas. “Our crop is so labor intensive that we as an industry are going to have to be leaders for this. We are actively engaged in all areas that reduce our dependence on labor.”

If growers are cagey about which robotic efforts they are considering, they have even less to say about whether they are using different planting methods in which robotics might play a role.

For the past 10 years, Spain-based Agrobot has been working with its partner growers to develop its strawberry harvester. Videos of Agrobot’s harvester online show substrate environments, although Agrobot’s creator said it can work with different farming formats.

“In every part of the world, they grow in a different way,” said Agrobot’s Inventor Juan Bravo. “There are many different ways to grow strawberries and many different growers. Our harvester can work on any system, but it’s about the efficiency of the harvester. In a more narrow bed, it is more efficient.”

Harvest CROO tested a portion of its prototype on a rail system at G&D Farms in Duette, Florida, in March. The rail system simulates how the harvester will maneuver through existing strawberry fields. Although this prototype has one picking head, the tractor, which is expected to have field time this spring, incorporates 16 picking heads and GPS technology for autonomous movement.

Agrobot’s latest model makes use of 60 picking arms.

“We can have a big quantity of low cost robotic arms to make this profitable,” he said. “The more narrow the bed, the more efficient the harvester can be.”

Agrobot counts strawberry grower Driscoll’s of Watsonville, California, as a partner testing its machinery.

Bravo declined to comment about the price tag associated with his harvester and whether growers would have the option to lease it, although in 2015 the price for Agrobot harvesters was estimated to be six figures.

“If you need the technology daily, it is usually cheaper to own it and that is why we prefer to sell the equipment,” said Bravo, referring to markets with longer strawberry seasons. In California, the harvesting season can extend from January through September.

Dollars and cents

Financially speaking, the math associated with using the Harvest CROO harvesting service, which is intended to reduce harvesting cost, is promising. The harvester is expected to do the work of about 30 human workers.

Harvest CROO plans to contract with growers in a manner similar to the piece rates growers pay to field workers. In return for the piece rates, growers will have use of the robotic machinery, which Harvest CROO will provide and maintain.

The prototype is able to identify, select and pick only ripe strawberries. The use of this technology is expected to reduce cost, energy usage and increase strawberry yields by 10 percent.

“These machines will cost less than human labor, and it will make competition irrelevant when you have something so much more reliable,” Wishnatzki said. “The premium growers are paying for H2A workers, at $3 and $4 a box, is not sustainable. Growers are doing it because they see that as their only guaranteed way to get the crop picked.”

Wishnatzki offered these ballpark estimates to put cost benefits into perspective: It can cost $10,000 to harvest an acre of winter strawberries in Florida annually and as much as $30,000 to do the same in California. He estimated a savings of at least 10 percent for labor costs, and he expects that to increase as the cost for the diminishing workforce continues to rise. He said costs are greater in California based on higher average yields.

“The machines will be modular so if there’s a failure on one part it is easy to replace and we can keep them harvesting,” Wishnatzki said.

This robotic strawberry picker, by Agrobot of Spain, uses 16 robotic arms and an artificial vision system to determine which strawberries to pick in this bed in Australia. Agrobot Inventor Juan Bravo has partnered with strawberry growers, including Driscoll’s of California, to develop the system.

How it works

Videos of Harvest CROO’s harvester in action reveal a picking head assembly with a wheel similar in movement to a Ferris wheel. With claws attached, it nimbly circles each plant to pick berries. The technology, known as the Pitzer Wheel, is the brainchild of Harvest CROO’s Chief Technical Officer Bob Pitzer. The harvester has 16 picking heads in action as it maneuvers through a strawberry field.

The picking apparatus does not act blindly. Cameras with imaging technology examine berries for quality based on color, mass and size. The harvester routes berries to fresh packs or juice trays, or discards them if needed.

Historically, hand-picked strawberries have been packed with extra fruit without the benefit of weighing technology in the field, but the harvester will be able to calculate how much goes into each consumer unit, giving growers an increase in yields and profitability.

The harvester’s mobile platform allows it to work autonomously in existing strawberry fields, which has been a top priority. Wishnatzki said the harvester can turn at the end of rows without extra space and crab walk to the next rows.

Future features for Harvest CROO’s robot include the ability to forecast production and scout for pests and disease. With images of the strawberries taken every three days, these will be associated with GPS locations giving growers new information for forecasting and intervention if needed.

Ready and waiting

Harvest CROO Investor Sam Astin III, a third-generation farmer in Wimauma, Florida, said he is ready for technology to transform his operation. New construction in his area has made harvesting labor scarce.

“Generally we’ve been able to get our berries harvested, but later in the season the price gets low and if you can cut back on your labor cost you can harvest more of your crop,” said Astin, who plants about 1,150 acres in winter strawberries. “It will make our life a lot easier. Having a backup plan – it’s a better deal than what we are doing now.”

Astin and other Harvest CROO investors have high praise for Wishnatzki and Pitzer, but Wishnatzki is not hanging his hat on it.

“We know growers cannot wait for this technology to come out,” he said. “These are competitors and the reason they have invested is not because they like me. It’s because they want early access to this technology to get these machines on these farms when it becomes available.”