Croft’s GrowPilot AI Links Five Japanese Tomato Greenhouses Under One Dashboard

Commercial tomato greenhouses run on climate computers, grower judgment, and a lot of walking between houses. On October 1, 2026, HortiDaily reported that Croft Agricultural Corporation—the Korean company behind the GrowPilot autonomous greenhouse platform—has deepened a partnership with Asai Nursery in Japan to bring crop-responsive AI into five tomato greenhouses under one management view. The memorandum of understanding marks Croft’s first move into Japan and its first strategic partnership with a large commercial grower outside Korea.

For readers who follow greenhouse tech from a gardening or CEA (controlled environment agriculture) angle, the story is not about replacing people with robots. It is about folding plant-level data into decisions that experienced growers already make—and testing whether five separate tomato houses can share a single dashboard without ripping out the climate system they already trust.

Who Croft and Asai are

Croft was founded in 2022 by CEO Hee Kyung Ryoo and CTO Woo Ram Lee. According to HortiDaily’s report, the company has already worked with more than 30 Korean growers. Its platform centers on GrowPilot for autonomous greenhouse control and GrowBot for scanning and phenotyping plants as they grow—not only logging the air and irrigation around them.

Asai Nursery brings a different kind of scale. HortiDaily describes a 119-year, fifth-generation family operation based in Tsu City, Mie Prefecture, led by CEO Yuichiro Asai. The nursery has grown tomatoes in Mie for the last 18 years. It runs five tomato greenhouses across 13 hectares and produces more than 1,000 tons of cherry tomatoes a year, plus two kiwi orchards across 15 hectares. The company is expanding into a new greenhouse in Hokkaido and running early-stage avocado research in Okinawa, with exploratory work in wasabi, Japanese pear, and strawberries. HortiDaily quotes Yuichiro Asai’s framing of the business as “a production business anchored in R&D.”

That R&D posture already shows in Asai’s existing toolkit. The nurseries run Priva for climate control and data. Since 2018, Asai has also held a joint venture with DENSO Corporation called AgriD, developing large-scale, next-generation greenhouse technology in Mie Prefecture. Croft is arriving not as a blank-slate installer but as a partner plugging AI into a house that already knows sophisticated automation.

What GrowBot and GrowPilot actually do together

At the center of the partnership is GrowBot, Croft’s system for scanning crops and phenotyping each plant as it grows. HortiDaily emphasizes the distinction: many greenhouse platforms track environment first; GrowBot is built to read the plant directly and feed that phenotyping data into GrowPilot, which turns the stream into environmental decisions.

GrowPilot’s other design claim matters for a Priva house. HortiDaily reports that the platform is sensor- and environment-agnostic—it can draw data from whatever hardware a greenhouse already has installed. For an operation Asai’s size, that removes the usual adoption tax: replacing working climate computers and sensors just to try a new software layer. Integration of Croft’s AI into Asai’s Priva-based systems is progressing as Asai finishes ongoing greenhouse renovations, the report says. Site surveys are complete at both companies’ facilities, with representatives visiting each other’s operations. Installations of new proprietary sensors Croft is developing are planned for October.

In plain terms, the stack looks like this: existing Priva hardware stays; GrowBot adds plant phenotyping; GrowPilot consolidates decisions; new Croft sensors fill gaps during the October install window. The partnership aims to consolidate all five of Asai’s tomato greenhouses under a single grower working from a single dashboard, rather than the industry pattern of closely monitoring each facility by hand.

Labor, expertise, and what “efficiency” means here

Croft and Asai are targeting yield increase and quality improvement, along with less time spent on day-to-day decisions such as crop registration, HortiDaily reports. The labor angle is carefully framed. Commercial greenhouse operators everywhere are short on experienced growers. Croft’s approach, as described in the article, is to reduce greenhouse labor dependency not by cutting positions but by encoding expert-level decision-making into the system so growing knowledge is not trapped in one person’s memory. Scaling five large commercial facilities across 13 hectares without a proportional jump in specialized staff is the test the partners are running together.

A Korean reference point appears in the same HortiDaily piece. JE Farms, a cherry tomato grower in North Jeolla, South Korea, saw a 45 percent gain in operational efficiency after adopting Croft’s systems without hiring more staff for crop registration—company-cited results from Croft’s existing Korean deployments, not an independent audit of Asai’s future houses. The Japan partnership is presented as the next real test of integrating that crop-responsive approach into day-to-day farming at a different scale and culture.

An equal partnership, not a bolt-on sale

HortiDaily describes the MOU as an equal partnership rather than a one-way client install. Asai contributes the commercial greenhouse environment, decades of grower expertise, and infrastructure already in place. Croft contributes GrowPilot and automation that can work across environments and sensor sets. That framing matches how Asai already behaves—Priva, AgriD with DENSO, multi-region crop experiments—as a grower that treats technology as continuous R&D rather than a single purchase order.

Hee Kyung Ryoo, Croft’s CEO and co-founder, is quoted in HortiDaily on the collaboration: “This is a very collaborative arrangement, and Asai has been the perfect partner for us to work with in expanding the capabilities of our management system.” On Japan as a proving ground, Ryoo adds: “Japan is one of the most technologically advanced markets in the world for controlled environment agriculture, with one of the largest and most sophisticated CEA sectors anywhere. That combination is exactly where we want to prove this works.” The article also carries Ryoo’s longer view that the partnership will be “one of the biggest steps on our journey to help commercial greenhouse operators worldwide find ways to expand controlled environment agriculture without being bottlenecked by headcount.”

Those quotes matter because they set expectations. Croft is not claiming Asai’s five houses have already hit a published yield percentage from GrowPilot. The October story is about surveys done, sensors planned, Priva integration underway, and a dashboard goal across 13 hectares of tomatoes.

Why gardeners and small growers should care

Home gardeners will not install GrowBot over a San Tan Valley raised bed. The transferable idea is narrower and more useful. Plant-responsive growing—watching the crop, not only the thermostat—is the same habit that separates a healthy cool-season bed from a schedule followed blindly. Consolidating information so decisions do not live only in one person’s head is how community gardens and small farms survive staff turnover. Working with existing hardware instead of demanding a full rip-and-replace is how technology actually enters working greenhouses.

Japan’s CEA sector is a high bar. If crop-responsive AI can sit beside Priva in a fifth-generation cherry tomato operation producing more than a thousand tons a year, the lesson for the wider industry is practical: phenotyping plus an agnostic control layer may matter more than a brand-new climate computer. For Croft, founded only in 2022 and already past thirty Korean growers, Asai is the first large commercial proof point outside Korea. For Asai, the partnership extends a long habit of pairing family production with outside engineering—DENSO yesterday, Croft’s GrowPilot and GrowBot today.

HortiDaily’s October 1 report leaves the story where commercial agriculture stories often begin: an MOU, site surveys finished, sensors slated for October, renovations still in progress, and a clear operational target—five tomato greenhouses, one dashboard, plant data in the loop. Yield and quality gains remain goals, not published Asai results. The JE Farms 45 percent operational-efficiency figure remains a Korean reference, not a Japanese guarantee. What is solid in the reporting is the partnership structure, the technology roles, the Priva compatibility claim, and the quotes from Croft’s CEO about collaboration and Japan as a proving ground.

For GTH STV NEWS readers who track AI in food growing, that is enough to watch the next season of Asai’s cherry tomatoes with a sharper question: when five houses share one screen, does the crop still lead the decisions—or does the dashboard start leading the crop? Croft and Asai say they are building for the first answer. October’s sensor installs will start testing whether the plants agree.