Israeli startup Picker AgRobotics says its flexible-arm harvesters will pick a 900-pound citrus bin for $20 in California's Central Valley, against roughly $25 for manual crews — a 20% cost cut from day one for growers who lease the machines.

The company, founded in Israel in February 2025, has built four prototype units and plans to launch customer pilots in January 2027, starting in Israel, then California, then probably Spain, according to cofounder and CEO Yossi Rosenblum, who spoke at the World Agri-Tech innovation summit in London.

How does the technology differ from other robotic harvesters?

Instead of the rigid pick-and-place arms most orchard robotics companies deploy — arms that reach into a tree, grab a single fruit, retract and place it on a conveyor — Picker AgRobotics uses long, hollow, flexible arms that snake through branches. The system, branded "Snip-N-Roll," cuts the stem and lets the fruit roll through the arm directly into the grower's existing bin.

"All of the companies we've seen, they pick and place," Rosenblum said. "They go all the way into a tree, pick one fruit, then go all the way back, and place it on a conveyor. That is a very expensive machine and not very efficient."

Rosenblum said the design avoids grippers entirely: "There's no gripper. We just approach the fruit, cut the stem, and let the fruit roll through the arm." Because the arm is a pipe, the machine does not require growers to change orchard architecture — no high-density planting, trellising or special pruning.

The company holds patents on the flexible and hollow robotic arm designs, which Rosenblum described as the first of their kind for tree harvesting. Flexible robotics has gained ground in other industries, he noted, but "has done nothing in agriculture so far."

What are the coverage and cost targets?

Picker AgRobotics trained its systems on avocados — large, untrimmed, deep-canopied trees with green fruit against green leaves — on the theory that mastering avocados makes citrus, pears and apples straightforward. The company aims for 95% fruit recovery, versus roughly 30% for the short rigid arms now available, Rosenblum said.

The economics rest on light industrial materials. Extending the arm by a meter or two adds little cost; the expense sits in the motors and computer, which the company has already priced. Leased machines are designed to undercut manual harvest costs by 20%, and purchased units by 50% over time.

The units are towable rather than fully autonomous individually. One tractor operator — or one autonomous vehicle — could move four or five units between trees, avoiding the cost of outfitting each machine with its own navigation stack.

What data does the harvester generate?

The machines carry AI vision models that make per-fruit decisions on ripeness, size and quality, letting growers harvest what sells while logging tree-level data on what was picked and what remains.

"So far, you've been picking blind. You have no idea what fruit is coming from which tree, which tree has a problem, maybe there's a deficiency," Rosenblum said. "Now you're going to pick, and it's not going to be blind."

How will it reach growers?

Picker AgRobotics will lease first to lower adoption risk — "You cannot tell the farmer who doesn't know anything about this machine: just put down the money," Rosenblum said — and sell machines with service and upgrade contracts once demand scales. The company will sell direct to very large growers, several of which it is already in contact with, and use existing dealers and distribution partners for mid-sized and small operations.

The 2027 pilot deployment will add five more units for a fleet of six, with teams staying on-site with customers until performance is validated before wider rollout.