University of Granada launches AI system to certify sunflower oil quality

Researchers at the University of Granada have created an artificial‑intelligence model that rapidly quantifies oleic‑acid content in commercial sunflower oil and classifies samples as conventional, medium‑oleic or high‑oleic. The method combines Raman spectroscopy with spatial offset (SORS) and chemometric analysis, delivering results in under 120 seconds without chemical reagents.

Validated on more than 300 blended samples, the AI‑based tool matches the official gas‑chromatography reference and is intended as a fast screening technique for producers, packagers, internal laboratories and regulatory bodies. It aims to detect labeling mistakes, undeclared blends or fraud, especially under Spain’s new Real Decreto 351/2025, which governs optional sunflower‑oil labeling based on oleic‑acid levels. Lead researcher Guillermo Jiménez Hernández emphasizes that the approach complements, rather than replaces, official testing.

The system could lower verification costs, speed up quality checks on production lines and improve traceability throughout the agri‑food chain.