b'standardized quality data across labs glob- over the germination process fromMILESTONE. WHAT DOES THIS REC-ally which ensures stable quality results inseed to plantOGNITION MEAN FOR PHENOLYTICS, the fieldat a high-throughput and with theAND HOW MIGHT IT SHAPE YOUR Lack of automated solutions topossibility of full automationNEXT STEPS (E.G. R&D, COMMERCIAL create high-throughput data to leverageWith this, we can create real biomassEXPANSION, COLLABORATIONS)? AI and big data for R&D and breeding.ground-truth data rather than indirectJW: This is indeed a big validation for us Fully-automatable, high-throughput dataproxies. This ground-truth biomass datathat comes just at the right time. While generation of seeds and seedlings as part ofis crucial to close the genotype-phenotypewinning the InnovAction Stage in 2023 was routine testing, allowing to build predictivegap, and you always need the ground-truthan important validation for us personally seed-seedling models at ten-thousands ofto make the most of proxiesthat we are indeed on the right track, this datapairs per day, and accelerating R&DThis makes this technology so robustyears Euroseeds win should act as a valida-and breeding with actionable big data fromand universally applicable: provided thattion to the industry that these technologies the lab differences show up in density, volume,are market-ready and worth investing in.Long lead times because every germi- shape or structure and that we can resolveWhat we have seen over the last few nation test adds weeks to each step, e.g., inthem (which we usually can with a reso- years is that across the industry, there is a production & processing, preventing fastlution in the low m range), we can alsogreat interest and recognition of the poten-decision-making. Quick, few minute seedmeasure them. With this, we can phenotypetial of our technologies. However, the tran-bulk scans to predict usable seeds directlyseeds and seedlings at a high throughput insition from interest to widespread adoption after harvest, detect quality problems earlyautomated workstreams, extract measure- has been a challenge (as with most disrup-and steer seed processing with quick bulkment data and classify with AI and directlytive technologies), and while many want to scans in-line, to integrate seed quality datalink this data output to clear decisions forbe the pioneers, few want to take the risk across the supply chain and enable earliereach seed lot. With the new phenoMatchto be the first.decision makingsolution, we can also 1-1 phenotype seedsSo, we have spent the last years further High waste in seed processing dueand seedlings at a high throughput, linkingdeveloping, validating and thus derisking to an inability to diagnose problemsand predicting germination and seedlingthe technologies for ever broader appli-and inability to sort for seed quality andquality/vigour to seed quality with everycations, so that we now finally can confi-germination outcomes on a single seedroutine quality test.dently say: we are not just market-ready, but level. Quickly characterize each seed lotWith the phenoScort, we also havealready in the market with a good pipeline and seed, set the right sorting parametersthe possibility to translate this diagnosticof system implementations ahead. and tailor the border between usable andseed-seedling data directly into AI-basedAs the adoption is now picking up non-usable seeds to prevent valuable seedssorting decisions based on predictiveand the general interest translating into and lots from going to waste. Sort directlymodels, to sort not just for 3D seed quality,real urgency to adopt the technologies to for seed quality, and with predictive modelsbut also seedling quality. unlock value with them, the Euroseeds win for seedling quality, on a single seed levelis a strong catalyst to accelerate this process. at a high throughput with the phenoScort. SWE: WINNING FIRST PRIZE AT THEIn parallel to the validation that comes with We thus enable the creation of ground- 2025EUROSEEDSINNOVACTIONnew proven use cases and real-life imple-truth data which can be shared across func- STAGE MUST HAVE BEEN A MAJORmentation, the win acts as a further signal tions and across organizations to steer and optimize all operational processes and pro-vide breeding and R&D with the actionable big data needed to accelerate discovery and innovation. SWE: WHAT MAKES PHENOLYTICS APPROACH PARTICULARLY INNO-VATIVE COMPARED WITH EXISTING METHODS (E.G. MANUAL SEED TEST-ING, TRADITIONAL 2D IMAGING, VISUAL INSPECTION, OR CONVEN-TIONAL BREEDING/QUALITY CON-TROL)? JW: 3D X-ray Computed Tomography (CT) is the only energy source known in physics that can tick all of the following boxes: Measure internal & hidden struc-tures (e.g., seed organs or roots in medium) in 3D as volumes and real biomass3D X-ray Computed Tomography shows the seed quality differences in sunflower and melon seeds.data FEBRUARY 2026|SEEDWORLD.COM/EUROPEISEED WORLD EUROPE I 7'