Artificial intelligence is being used to protect one of the wine industry's oldest traditions: the use of natural cork closures. Portuguese cork producer M.A. Silva has introduced an AI-powered inspection system designed to improve the quality, consistency, and traceability of corks used in wine bottles. The initiative demonstrates how modern AI can strengthen a centuries-old craft without changing the natural production process itself.
The AI system performs multiple independent inspections on each cork, analyzing it from different angles and across numerous quality parameters. Before a Nobeltech cork is classified, it undergoes 12 separate inspections that assess factors such as TCA contamination (the compound responsible for cork taint), mechanical performance, sealing ability, structural cracks, clay contamination, lenticel exposure, and even insect holes. Unlike human inspectors, the system is not affected by fatigue and can maintain near-perfect consistency across large production volumes.
Beyond detecting defects, the technology is helping improve the long-term performance of corks. AI analyzes historical data to predict oxygen transmission rates (OTR), which influence how wine ages in the bottle. More consistent OTR across a batch means wines are more likely to develop uniformly over time, preserving their intended taste profile and quality. M.A. Silva's research indicates that AI-selected corks show highly consistent OTR after one year, along with consistent wine taste profiles.
The system also provides full traceability at industrial scale. M.A. Silva can produce up to 40,000 corks per hour, and every inspection is recorded with imaging and classification data. This allows winemakers to verify the quality decisions made for each individual cork. Importantly, the article emphasizes that AI does not replace the traditional cork itself; rather, it enhances quality assurance, helping preserve a natural and sustainable practice that has been used in winemaking for thousands of years.
