AI-Driven Sensing for Predictive Wine Quality and Enogastronomic Experience

Objective

The FLAVOURID project proposes an innovative, low-cost and miniaturized sensing solution that combines chemical, sensory and emotional profiling with smart technologies and artificial intelligence (AI) to monitor wine quality in real time and enhance personalised food pairing experiences. The non-invasive sensing device will monitor colour, aroma, off-flavours, acidity and other sensory-relevant parameters throughout bottle aging, enabling dynamic profiling without opening the bottle. Sensor data will be processed in real time and linked to an AI-powered mobile application. This app will offer users insights into (1) real-time wine profile (sensory and chemical profile and prediction of optimal temperature of consumption), (2) wine aging monitoring information (detecting off-flavours and spoilage markers as well as predicting ideal time of consumption), (3) personalized food pairing recommendations based on integrated chemical, sensory, and emotional data. The AI model will be trained using a robust dataset collected from expert sensory panels, chemical analysis (e.g. FTIR, HPLC, GC-MS), and consumers' emotional responses captured via FaceReader® software. The work plan is structured in six work packages: WP1 (project management), WP2 (wine sample collection and characterization), WP3 (smart device development), WP4 (AI and app deployment and validation), WP5 (training and career development), and WP6 (dissemination, exploitation and communication). A secondment at Foodpairing® (Belgium), a pioneer in AI-based flavour science, will support the co-development and validation of the AI-driven pairing model in real gastronomic scenarios. FLAVOURID offers a scalable (spin-off), patentable solution aligned with EU goals for digitalisation, sustainability and food innovation, while advancing the researcher’s expertise in AI, sensing, and interdisciplinary food systems.