top of page

Running a UV-C trial in commercial strawberries, with Sant'Orsola, Antobot and Croptiq

Oct 28
5 min read

What does it take to run a serious UV-C trial in a commercial strawberry crop?

The answer is less exciting than it should be. Robots are one piece. The grower, the pathology team, the technology provider and the treatment strategy all have to work in parallel, or the result is noise. The Sant'Orsola trial we completed in September 2026 is a good illustration of that, and it produced a strong commercial signal in favour of a reduced chemical programme supported by UV-C. The headline figures and the full statistical analysis are documented in the official scientific report prepared by Sant'Orsola within the Bioconfarm project. That report is available on request.

The purpose of this blog is different. I want to describe what was done, by whom, and why each contribution mattered.

The setup

The trial ran from late July to the end of September 2026 in the framework of the Bioconfarm project funded by the European Union. Sant'Orsola Cooperative acted as a Living Farm Lab at one of its associated commercial strawberry farms. 17 adjacent high tunnels, cv. Falco, four plants per pot on peat substrate. Ten tunnels followed the conventional chemical programme. Seven central tunnels followed a reduced chemical programme integrated with UV-C.

The question was clear. Can a reduced chemical programme supported by UV-C maintain effective powdery mildew control in a commercial strawberry crop, without pushing the plant into stress?


Who did what


Sant'Orsola Cooperative created the Living Farm Lab environment, coordinated the trial design, and ran the plant pathology discipline across the full assessment period. The pathology team assessed powdery mildew on six dates, examining 320 leaves per assessment, and separately tracked spider mite incidence, thrips-related fruit damage and leaf chlorophyll status. The Sant'Orsola team also operated the UV-C robot on site each night, since the platform is semi-automatic and the current tunnel infrastructure (uneven rows and gaps between tunnel sections) does not yet support full autonomous operation. Without that rigorous evaluation and the hands-on operational commitment, there is no credible science behind any claim made from this project.


The commercial grower opened a real production crop to the trial and adapted normal crop-protection practice around the different programmes while keeping the operation running under commercial constraints. That is a non-trivial contribution. A grower who runs two different crop-protection strategies in parallel, in a crop that pays their bills, is accepting real commercial risk in exchange for a result that benefits the wider industry. Without the grower, the trial does not exist.


Antobot supplied the UV-C robotic platform and the application equipment, provided technical support throughout the trial, and remained the reference point for any question on the technology. The platform is honest about where it is on the autonomy curve: fully autonomous on compatible infrastructure, operator-run where the site does not yet allow it. At Sant'Orsola it ran at night, from 9pm, across eight weeks of UV-C treatment. Without a reliable machine executing the programme as designed, there is no treatment to compare.


Croptiq designed the UV-C treatment strategy. That included the dose progression, the robot speed at each phase, the application frequency, the plant acclimation window after planting, the night-time scheduling, in-canopy dose verification at three plant positions, and the integration of UV-C with the reduced chemical programme as the crop developed. The protocol was adjusted in step with the crop. Without that strategy, UV-C is a machine looking for a protocol.

Remove any of these four contributions and the trial produces a different outcome, or no outcome at all.


The treatment approach

Plants were transplanted on 25 July. UV-C treatment began two weeks after planting, on 10 August, after the crop had been established and had received two conventional fungicide applications in common with the chemical strategy.

For the first two weeks of UV-C treatment, the UV-C tunnels received one application per week at 2 km/h, corresponding to approximately 75 J/m² per application. The 75 J/m² figure was estimated from irradiance measurements at three canopy positions: above the canopy, beneath the canopy, and at the plant crown.

Once the plants had acclimated, the programme stepped up to two treatments per week at 1 km/h, approximately 150 J/m² per application. All UV-C applications were made at night, starting at 9pm. The night-time choice matters more than it sounds. Visible and UV-A light activate the DNA repair pathways that pathogens use to survive UV-C damage, so day-time or dawn applications lose part of their effect. Night scheduling helps preserve UV-C effectiveness by limiting light-driven pathogen DNA repair.


The chemical programme shift

In parallel, the chemical programme in the UV-C tunnels was progressively reduced. Conventional fungicide approximately every five days during the acclimation phase. Then approximately every fifteen days. At the start of harvest, conventional fungicide was stopped in the UV-C programme. From that point the programme ran on UV-C and lupin seed extract (Problad, Certis Belchim). The last treatment was applied on 28 September.

This was not UV-C replacing every chemical. It was UV-C as the central tool in a different integrated crop-protection strategy.


What happened in the crop?

By 30 September, powdery mildew incidence was 79.4% in the conventional chemical strategy and 1.9% in the reduced-chemical + UV-C strategy. Disease severity was 17.20% and 0.38%, respectively. The divergence over time is the most important outcome of the trial. These figures describe the two integrated management strategies; they do not isolate the effect of UV-C alone.



Figure 1. Powdery mildew incidence over time. Original Figure 2, Sant’Orsola / Bioconfarm report (2026). Reproduced subject to permission.

Honest limitations

The two strategies were allocated by tunnel position, not by randomization. The Sant'Orsola report treats the statistical comparisons as exploratory associations rather than definitive causal proof. That framing is important and belongs in any serious discussion of the results, which are documented in the official report.

A secondary observation worth flagging

The trial also looked at spider mites and at thrips-related fruit damage as secondary outcomes. On spider mites, the two programmes were statistically indistinguishable, which is consistent with UV-C not being an acaricide.

The thrips observation was less expected. UV-C is not known as an insecticide against thrips, and the starting assumption was that the two programmes would show similar damage levels. Averaged across the two assessment dates, thrips-related fruit damage severity was significantly lower in the UV-C programme than in the conventional chemical programme (0.59 versus 1.36 on a 0 to 5 ordinal scale, p = 0.0286). At individual assessment dates the differences did not reach statistical significance, which the report attributes to between-tunnel variability.



Figure 2. Mean thrips-related fruit damage (0–5 scale), averaged across two assessment dates. Original Figure 6, Sant’Orsola / Bioconfarm report (2026). Exploratory result; reproduced subject to permission.


The mechanism behind this signal is not established. Candidates worth testing include an indirect effect through reduced spray disturbance of natural enemy populations, a change in canopy microclimate, or a direct effect on thrips behaviour that has not been previously documented in strawberry UV-C work. For now this sits in the exploratory category and belongs in the "worth investigating under a controlled design" column, not in the "proven" column.

More broadly, the thrips signal is a reminder that UV-C in a protected crop can produce a wider set of secondary effects than disease suppression alone. Plant defence induction, changes in surface chemistry and volatile profiles, impact on natural enemies, and others all deserve systematic characterisation in future work.


What I take from this

A UV-C robot does not reduce chemical use on its own. The grower does. The robot is one tool in a strategy that must be designed, implemented day to day, monitored and verified in the crop.

The Sant'Orsola trial showed what becomes possible when growers, researchers, technology providers and crop-protection specialists work together. The promising results are not simply about using UV-C. They are about applying the right technology, with the right strategy, under real growing conditions.


That is what a serious collaborative AgTech project looks like and that is why we work with

through the complete process:


Assess. Select. Deploy. Optimize. Verify.



 
 
 

Comments

Rated 0 out of 5 stars.
No ratings yet

Add a rating
bottom of page