Moët Hennessy taps AI to sniff out a hidden wine defect
Moët Hennessy, LVMH's fine-wine division, worked with Analog Devices (ADI) and UC Davis on a system combining chemical sensors with machine learning to flag a defect called Fresh Mushroom Aroma (FMA) while wine is still juice. The team says the "scent fingerprint" model identified high-risk samples ahead of conventional methods with 99% accuracy. FMA does not appear every year — the most recent notable occurrence was the 2023 harvest — and typically only shows up after fermentation, when losses can reach millions of euros.
It is a concrete example of AI plus chemical sensing moving food and beverage quality control from after-the-fact salvage toward early prevention.
酩悅軒尼詩（Moët Hennessy）正將人工智慧與化學感測技術導入葡萄酒釀造流程，試圖在發酵問題發生前，預先找出可能影響酒品品質的隱形風險。這項由 LVMH 旗下高級葡萄酒部門、亞德諾半導體（ADI）與加州大學戴維斯分校（UC Davis）共同推動的研究，聚焦於一種名為「新鮮蘑菇香氣」（Fresh Mushroom Aroma，FMA）的缺陷，目標是在葡萄汁階段就提前辨識異常，而非等到發酵後才發現。
根據研究團隊表示，FMA 近年已在香檳產區等地出現，且並非每年都會發生；最近一次明顯出現是在 2023 年的採收季。這種缺陷在葡萄汁階段幾乎無法透過人類感官察覺，直到進入發酵階段成為葡萄酒後才會顯現，屆時往往已造成難以挽回的損失，甚至可能導致數百萬歐元的廢棄浪費。
#Moët Hennessy#LVMH#Analog Devices#UC Davis#Robert-Jean de Vogué Research Center