Renji John gets a second shot at greenhouse robots with Eternal.ag
The Cologne startup recently raised about $10 million to scale autonomous tomato harvesting, a hard robotics problem its CEO has already seen fail once.
By Ryan Merket ยท Published
Why it matters
Eternal.ag is a founder comeback story inside a hard robotics category: if John and Kunjachan can make tomato harvesting reliable, greenhouse automation moves from demo to budget line.
Renji John is scaling Eternal.ag's autonomous tomato harvester after the Cologne agritech startup recently raised about $10 million, per Forbes, turning his second run at greenhouse robotics into a live test of whether physical AI can survive commercial greenhouse work.
Forbes revisited Eternal.ag on June 21 in a broader look at physical AI moving into controlled-environment agriculture, reporting that the team has 26 employees split between Cologne and Bengaluru and is targeting one of greenhouse farming's least glamorous constraints: the availability of people willing to do repetitive crop work in hot, humid facilities. Eternal.ag lists Simon Capital, Oyster Bay Venture Capital, EquityPitcher Ventures, and Backbone Ventures among its backers.
John knows the trap in this market. He previously founded Honest AgTech in the Netherlands, another greenhouse robotics company. Forbes notes the business went bankrupt in 2023. That is the relevant founder context for Eternal.ag: this is not a first-time founder discovering the greenhouse labor problem from a pitch deck. It is a founder returning to the same operating environment after the first attempt failed.
Eternal.ag positions its system as purpose-built for greenhouses and designed to integrate with existing infrastructure. The company says its AI continuously learns to improve performance.
The wedge is labor, not a fully automated farm
Eternal.ag's first commercial product is Harvester, a tomato truss harvesting robot designed for high-tech greenhouses. Eternal.ag says Harvester can run up to 22 hours a day and requires no human operator. Those are company claims, not independently audited operating metrics.
The narrower scope matters. Forbes reports that the robot is shown working in a hydroponic tomato greenhouse, where plants are grown without soil, supported in media such as coconut coir, and fed through a water-nutrient mix. Seedlings still come from local nurseries, and planting remains human-operated. Eternal.ag is not selling the idea that every greenhouse job disappears at once. Eternal.ag is starting with harvesting, one of the repetitive, labor-sensitive jobs where uptime and cut consistency translate most directly into grower economics.
Eternal.ag also offers an ROI calculator for growers evaluating autonomy.
Why investors funded a difficult category
Agricultural robotics has repeatedly punished teams that confuse a controlled demo with commercial reliability. Greenhouses look structured compared with open fields, but they are still messy robotics environments: crops change shape and density, rows differ, humidity and corrosion punish hardware, and harvest quality has to be high enough that growers do not lose the crop value they were trying to protect.
That is why John's Honest AgTech history cuts both ways. It gives Eternal.ag a founder who has already seen the operational and financial failure modes. It also means Eternal.ag has to convince growers and investors that the second company is not simply a recapitalized version of the first. The answer Eternal.ag is offering is discipline: begin with tomato harvesting, retrofit into existing greenhouse infrastructure, and expand to more crop types only after the initial system proves itself.
Public materials do not disclose valuation, a lead investor, revenue, unit economics, or total capital raised to date. Forbes characterized the latest raise as about $10 million.
For Eternal.ag's backers, the bet is that labor scarcity will force greenhouse operators to automate even if the broader controlled-environment agriculture sector is still wrestling with energy costs. The direction of the pressure is easier to verify than any single statistic: greenhouses need steady labor across a year-round production cycle, and the work is physically demanding.
The greenhouse market is growing, but not cheaply
Controlled-environment agriculture (CEA) has a compelling sustainability story and a complicated cost structure. Forbes writes that CEA produces less than 1% of global agricultural production but is important in crops such as tomatoes, cucumbers, leafy greens, herbs, cannabis, and horticulture. The advantages are familiar: higher yield per square meter, year-round production, reduced pesticide use, and much lower water use.
The operating costs are less convenient. Citing Agritecture, Forbes notes that CEA can consume about 10x more energy than traditional outdoor farming and about 10x less water per pound of produce. Energy remains a sector-level problem. Eternal.ag is attacking the bottleneck closest to daily operations: whether a grower can harvest on schedule when labor does not show up, costs more, or cannot be staffed reliably across shifts.
Forbes also cites Van Noord Growers, a Dutch greenhouse operator, as a deployment proof point for Eternal.ag's robots. The useful missing detail is the commercial shape of that deployment. Forbes does not establish how many robots are operating, when deployment began, whether Van Noord is a paying customer, or what measured labor savings and harvest quality data look like. Until those figures are public, the Van Noord example supports market interest, not yet repeatable economics.
The real test is not autonomy in a video
The next step is turning autonomous tomato harvesting into equipment that European greenhouse operators treat as dependable, not experimental. The greenhouse market does not need another robotics demo. It needs machines that can handle hour 20 as well as hour one, move through existing facilities without expensive redesign, and cut fruit consistently enough that growers trust the system during peak production.
John has raised enough capital to make that case in more greenhouses. What the company has not yet disclosed is the evidence that will decide the category: paid deployments at scale, audited uptime, per-robot throughput, maintenance cost, and payback period. In physical AI, those numbers matter more than the word autonomy. Eternal.ag's second attempt will be judged by whether the robot turns the hardest part of greenhouse labor into a predictable line item.