A stone-fruit harvester is being built for the part of the season when a ripe peach or nectarine stops being a crop on a branch and becomes a clock. The machine is meant to improve harvesting efficiency, a fairly plain ambition until the fruit is ready all at once and the picking crew, packing line and orchard weather are all asking for the same few days.
That machine is one piece of a larger experiment taking place at Reservoir Farms, where 22 agricultural technology startups are operating in Salinas and Santa Rosa. The sites give companies something a conference room cannot: dirt, crop rows and the inconvenient timing of commercial agriculture.
From Demonstration to Working Orchard
The focus is less on futuristic machinery for its own sake than on recovering production that farms are already equipped to grow but cannot always harvest, thin or manage efficiently. Danny Bernstein, CEO and managing partner of Reservoir and Reservoir Farms, has argued that technology can help reclaim lost crops and strengthen California’s food security.
For specialty crops, that distinction matters. A tool that works in a controlled demonstration but struggles with irregular fruit, dust, trellis geometry or a short picking window is still a prototype. The useful version is the one that can be serviced, scheduled and trusted beside the rest of the harvest operation.
California Becomes the Test Plot
John Deere is directing research resources toward California agriculture through its support of Reservoir and the startups working there. The partnership places a major equipment maker alongside younger companies trying to turn narrow problems—such as harvesting stone fruit—into machines that can function at farm scale. The company’s support for Reservoir is part of a broader technology partnership.
The work is also being tied to grower feedback rather than left entirely to the companies developing the tools. UC Agriculture and Natural Resources has opened UC ANR Connect, a commercialization program built to connect commercially ready technology with California crop production and grower input. The program is accepting applications for its current cohort.
That field-level approach has already shown up in the Salinas Valley, where a demonstration brought technology providers into vegetable production to show tools for weeding, thinning, harvesting, drone application and irrigation management. The event was designed for growers to watch equipment work in the field and offer direct feedback.
For a California farm, the question is therefore less whether automation exists than where it can earn its keep first. Stone fruit has the urgency of a compressed harvest; Salinas lettuce has a different set of problems, with repeated passes and tight specifications. The same machine will not answer both, and the farm will still be the place where the answer gets judged.