Semi-Autonomous vs. Fully Autonomous: Where the Real Savings Are
“Automated” gets used loosely in greenhouse tech. The distinction that actually matters is simple: does a person still need to operate or supervise the machine while it works, or not.
A semi-autonomous system does part of the task on its own but still needs a person present, driving it, supervising it, or stepping in when it hits a case it can't handle. It might navigate a row automatically or handle the easy cuts while a person manages the rest. The machine reduces effort. It doesn't remove the person, or the person's cost, from the operation.
That operator's cost is the part easy to miss when comparing quotes. Run the numbers on a representative 10-hectare tomato operation: at typical robot-to-operator ratios and shift patterns, you need roughly 4 operators a day to run a semi-autonomous fleet at scale. Over 5 years, that alone comes close to €1.9 million, on top of the robots' purchase price and service fees.
Add it up against the labor it's meant to replace, and semi-autonomous automation on this model saves an operation only around €430,000 over 5 years; a fraction of what the headline "robot" price suggests, because the labor cost didn't go away. It moved from harvesting to operating the machine that harvests.
A fully autonomous system performs the task itself, without a human operating or supervising it in real time. In practice, that means the machine has to do, on its own, everything a person's eyes and hands would otherwise do: read plant height, canopy density, truss position and ripeness through real-time computer vision, work out the stem angle and where the fruit is relative to the leaves, then execute a precise cut, all while navigating row changes and lane switches without anyone directing it.
Where our fully autonomous harvester robot helps
Our Harvester runs up to 22 hours a day with no operator present, at all, which means the operator line in the cost model above doesn't just shrink; it disappears. On the same 10-hectare model, removing operator costs entirely is the difference between automation that pays for itself slowly and automation that actually changes the economics of the harvest. Growers don't buy a robot outright. They lock in a harvesting price today, set against the labour cost they'd otherwise be paying, and that price holds.
That structure matters because of where labour costs are heading, not just where they sit today. At roughly 15% year-on-year growth, the per-truss equivalent of labour cost roughly doubles within 7 years. Locking in the harvest price now, while labour is still comparatively cheap, is what makes the business case work in the grower's favour at signing, before a single robot has even reached their site.
You can do the math yourself. Our ROI calculator runs a 5-year comparison, using your own greenhouse's area, labor costs and current staffing, so you can see the actual gap between semi-autonomous and fully autonomous costs for your operation, not a generic industry estimate. Put your greenhouse details into it, and we'll show you what semi-autonomous harvesting costs today, and the number it becomes with eternal.ag.