Orchard Robotics is pitching specialty crop growers a first-year return of three to 10 times their investment — not from yield gains, but from cutting labor and input costs and tightening supply chain planning.

The company's AI-powered camera system, roughly the size of a large tissue box and weighing about five pounds, mounts on tractors and other farm vehicles and captures millions of images as they move through orchards and vineyards. Software measures fruit counts, size, color, growth rates, canopy health, water stress, disease and yield — data covering billions of fruit and millions of trees.

"A million numbers in a spreadsheet is not a product," founder and CEO Charlie Wu said on stage at last week's Ruggedize ag robotics conference at Reservoir Farms in Salinas, California. "Data is only useful if a grower has a mechanism of taking an action on that data easily. Your average grower does not want to be, nor should they be, a data analyst."

From images to action

Because many farms lack the connectivity to upload terabytes of raw imagery, the system processes images on an onboard NVIDIA Jetson computer. The output feeds into Orchard's FruitScope platform, where growers can inspect conditions down to individual trees.

The simplest use case is fruit count. Orchard can tell a grower how many fruit sit on every tree, informing pruning and thinning decisions. Variation within blocks is often substantial: half the trees in a block can carry double the fruit load of the other half, which determines where crews should spend more or less time.

"We're not actually selling hardware. We're not even selling software. We're not even selling data," Wu said. "We're selling the outcome that a grower sees by using all of it effectively."

Action takes two routes. Some growers run precision sprayers and spreaders that vary applications automatically based on insights from the platform. Most, however, receive instructions through a mobile app that tells workers where to go and what to do — a necessity, Wu noted, because few growers currently own precision application equipment.

Yield forecasts at 95%+ accuracy

Supply chain planning is the second major value driver. Wu said the platform predicts yields with 95% or better accuracy, informing decisions on hiring, bin procurement, storage requirements and what sales teams can promise customers.

"The biggest ROI is through reduction on labor and inputs," Wu told AgFunderNews. "When we sell our product, we don't actually claim to boost yield. We pitch them on the labor savings, the input savings, and supply chain efficiencies, and we say: 'Hey, you and I both know that if you do all these things correctly, you're going to see an increase in yield.'"

Scan frequency varies widely. Some growers scan five times a year; others scan a single field 25 times. The average runs around ten scans per season.

Pricing and business model

Orchard charges a per-acre, per-year subscription covering the software platform, unlimited scans and hardware access. When the company releases a new camera version, subscribers get the upgrade at no charge, avoiding obsolete equipment or forced repurchases.

Operationally, growers choose between a self-serve model — mounting the camera on their own tractors during routine mowing or spraying passes — and a full-service option where Orchard technicians scan fields and deliver results the next day.

Crop expansion and international pull

Orchard started in apples and now works in wine grapes, table grapes, cherries, blueberries, almonds, pistachios and citrus. Pomegranates, coffee and strawberries are among the new crops in development. Wu said the AI models learn across crops, so each new crop and farm improves performance across the board.

The company currently has "hundreds of systems in the field" scanning tens of thousands of acres, with ambitions to reach hundreds of thousands. It has grown multiple times year over year for the past several years. Most customers are in the US, but Wu reports "some pretty tremendous pull internationally" and rapid expansion abroad.

The company's stated goal is to eventually service every specialty crop on the market, positioning its per-acre scanning model as a substitute for manual labor allocation and guesswork-driven input spending across permanent plantings.