AI & Strawberries? Yes, Really: Inside ARGENTIC’s Strawberry Pilot

22 October 2026

What does an AI research project have to do with strawberries? More than you might think. In ARGENTIC’s fifth use case, agentic AI meets one of Europe’s favourite fruits, with the goal of helping to grow smarter, more resilient strawberry varieties for a changing climate.

Why strawberries?

Many strawberry nurseries in Southern Europe depend on a single imported cultivar, Fortuna. Climate change, water scarcity and soil salinisation are turning that dependency into a growing risk. New varieties that cope better with heat, drought and salt could make a real difference, for growers and for all of us who enjoy the fruit.

That is where BERRYPLASMA comes in. This Greek SME in Western Greece runs an integrated breeding, nursery and soilless cultivation programme, with a decade of trials behind it. The knowledge is there, and so is plenty of monitoring data. But the data sit in separate tools, and decisions rely on expert experience. ARGENTIC wants to connect the dots.

Three stages, one goal

The use case follows the whole life of a new variety:

• Breeding: crossing parent plants and selecting the most promising new genotypes.
• Cultivation: growing the best new varieties under commercial conditions to test yield, fruit quality and resilience.
• Nursery: multiplying healthy young plants of the selected varieties, ready for growers to plant.

From a sensor reading to a better protocol

Here is the plan. Battery-powered sensors in the greenhouse and the open field will measure substrate moisture, EC and pH, air and leaf temperature and humidity, light, wind and irrigation flow. ARGENTIC agents running on an edge device at the site will turn those readings into explainable, variety-specific advice, step by step:

1. SenseSensors sample substrate, climate, leaf and irrigation data every 1–10 minutes.
2. DetectA data-quality check, then early stress detection per bed and cultivar.
3. RecommendAn irrigation or fertigation proposal, with the readings behind it and a plain-language rationale.
4. DecideThe nursery agronomist approves, adjusts or overrides it.
5. ActA simple task for the nursery team, confirmed the same working day.
6. LearnData, decisions and outcomes are logged, and researchers validate variety-specific protocols.

The AI advises. The agronomist decides.

This is what we mean by agentic AI that is autonomous but human-aligned. The agents never operate the irrigation valves themselves: they monitor, explain and propose, and every decision stays with BERRYPLASMA’s experts. Each decision is logged, so the knowledge grows season after season.

What if, before it happens

What would a heatwave do to a new variety? Or a dry spell, or saltier water? A digital twin will let researchers explore drought, salinity and heat scenarios for each cultivar before they happen in the field. Validated results will become versioned, reusable protocols in a knowledge graph, so every season builds on the last.

Breeding know-how stays at home

A new strawberry variety takes years of work, and that know-how is precious. That is why raw data will stay on site at BERRYPLASMA. Only anonymised, aggregated knowledge may leave, without cultivar identity or exact location, and nothing is released without BERRYPLASMA’s approval. Processing happens close to the plants, on an energy-aware edge set-up.

Where we are now

We are in month 5 and the use case is in its design phase. Requirements and the data management plan for UC5 are in place, and the scope, user roles and KPIs are defined. Deployment design and equipment procurement are under way. The first sensors and an initial dataset are planned for spring 2027, with full operational validation planned from December 2027.

One piece of a bigger puzzle

Strawberries are just one of ARGENTIC’s seven pilots. Across healthcare, energy, marketing, tourism, agriculture, recruitment and virtual companions, we are testing the same idea: trustworthy, energy-aware AI agents that work where data is created and keep people in charge.

So the next time you enjoy a strawberry, remember: the varieties of tomorrow may get a little help from AI, with people firmly in charge. 🍓

👉 Explore all ARGENTIC pilots