DUETTE – You may be eating robot-picked strawberries before you buy your first self-driving car.

“I think the two technologies are competing,” said Manoj Karkee, an associate professor of biological-systems engineering at Washington State University and an authority on automated agricultural systems. “Either one can be here sooner.”

While commercial self-driving vehicles, such as delivery trucks and cabs, might be in widespread use by the early 2020s, he added, consumer acceptance of the technology is probably farther off, maybe 10 years away.

The people at Harvest CROO Robotics in Duette certainly hope Karkee is right about a robotic strawberry picker, and they are racing to become the first company to cross that finish line.

“We hope to have a machine commercially available by 2020,” said Gary Wishnatzki, CEO of Wish Farms, a Plant City grower and packinghouse operator and a founding investor in Harvest CROO. “My fear is if we don’t solve this problem (of adequate harvest labor) with automation, this will be a much smaller industry. The price of strawberries will be higher and the farms smaller.”

Strawberry farmers in Florida and California, the largest producing state, are already struggling to find enough human labor to pick their fruit, said Wishnatzki and Michael Christensen, director of technical services and global research and development for Driscoll’s, the world’s largest berry grower.

“I think the industry is eagerly awaiting commercial solutions to its labor demands,” Christensen said. “It's about revenue preservation – if you can’t harvest the fruit, you can’t sell it.”

Driscoll’s is one of several investors in Harvest CROO, which also includes Sweet Life Farms and Astin Farms in Plant City and large California growers such as California Giant Inc. and Main Street Produce Inc., according to company information. Christensen was in Duette last week to see its progress toward a commercial machine.

“I think it’s impressive progress,” he said.

When asked how soon Driscoll’s could use a robotic picker, Christensen replied, “Today.”

Harvest CROO’s investors account for at least two thirds of annual U.S. strawberry production, Wishnatzki said.

The current robotic picker sits on a four-wheeled platform spanning six standard rows of strawberry beds and meanders at little more than 1 mph, said Bob Pitzer, Harvest CROOs chief technical officer and, like Wishnatzki, an original investor. It picks simultaneously over four beds.

As it moves along, a photographic sensor picks out a ripe strawberry, which then activates the “Pitzer wheel,” a patented, six-arm wheel each with a plastic clamp that gently grabs the berry. The machine puts the picked berry into a plastic container, which carries the fruit to an overhead packing system.

The strawberries are packed into a standard, one-pound plastic shell until another machine comes to pick up the load. The goal is to carry 1,000 pounds of picked berries until it needs to stop for unloading.

The first working prototype of the harvester is being tested on a 600-acre Wish Farms field in Duette, about 20 miles south-southwest of Mulberry, said Scott Jantz, a Harvest CROO electrical engineer.

“What we’ve proven is the basic idea of taking a large machine and putting it over strawberry plants,” he said. “The vision system and the robot picker itself is getting a chance in a real field.”

The company has proved both systems work and is developing further refinements, including increasing speed and increasing the time of continuous operation, Wishnatzki said. The goal for the latter is to run at least 15 hours per day.

Harvest CROO is already working on a second prototype that will be ready in the fall for the 2018-19 Florida strawberry season and will also be tested in California, Jantz and Pitzer said.

The second prototype will be narrower, with wheels traveling over two strawberry beds but still able to pick over four beds, they said.

The goal for the new prototype is to pick one berry per second, an improvement over the current rate of three berries in eight seconds, Pitzer said.

“As soon as we get something reasonably good, we won’t be able to build them fast enough,” he added.

Pitzer estimated it would take 2,000 robotic harvesters to pick 100 percent of current U.S. production.

Further into the future, Pitzer saw the development of robotic pickup and delivery systems that would dock with the harvester like to two orbital space vehicles hooking up, he said. That would enable the harvester to unload the picked strawberries and continue operating with little or no interruption.

Harvest CROO’s business model calls not for selling the machines to growers in Florida and California, the two major U.S. producing states, but to offer harvesting services through the company to those growers, Wishnatzki said. That’s already a $1 billion market.

Although it has the deep pockets, Driscoll’s is comfortable with contracting for robotic harvesting rather than buying its own machines, Christensen said.

Karkee, who is working on a robotic harvester for the apple industry, agreed that’s a more likely path for introduction of the machines, making it more likely strawberry and apple growers would adopt the new technology.

“It is easier, from the farmers’ perspective, to use if they don’t have to spend millions of dollars,” he said.

Although Karkee has not personally examined the Harvest CROO system and could not make a professional evaluation, he’s read enough to be impressed by its progress.

“This company could be ahead of the curve,” he said. “I’m pretty sure in another three to five years, there will be something that will work.”

The next barrier might be economic and not technological.

Wishnatzki argued that strawberry growers would be willing to pay more for a robotic harvester than they do for manual labor because they're currently leaving a lot of unpicked fruit the harvest in the field.

But Karkee said farmers would not adopt a mechanical system unless the company can show the cost of harvesting will not exceed the farm price they would get for selling their fruit or vegetable.

“If the cost is more than leaving your fruit in the field, they won’t use it,” Karkee said.

Some early adopters may be willing to use mechanical systems at a loss for a couple of years, Karkee added, but only if there’s a reasonable expectation that costs would fall below that break-even point eventually.

That’s not an issue for Driscoll’s, Christensen said.

“It’s not hard to imagine saving 20 to 25 percent off labor costs,” he said.

Kevin Bouffard can be reached at kevin.bouffard@theledger.com or at 863-401-6980.