Fermentation is a critical stage in the production of spirits and requires constant monitoring. Automated monitoring in the industrial sector is still focused primarily on temperature, leaving the full potential of advanced sensors and data analytics untapped.

The study by Horváth et al. analyses scientific solutions for automated monitoring, evaluating them according to sensor type, the nature of the mash, and the level of reproducibility and feasibility. The parameters monitored include temperature, pH, sugars, alcohol, CO₂ and aromatic compounds, using technologies ranging from traditional sensors (RTD, NDIR) to biosensors and e-Nose/e-Tongue devices, which can generate ‘fingerprints’ of the fermentation process. However, application complexities and operational challenges currently limit their industrial uptake.

There are few studies that genuinely focus on in-line and online monitoring; many solutions remain confined to the laboratory. Among the most relevant applications are ultrasonic systems for density, balances for CO₂ calculation, optical sensors for alcohol and colour, and biosensors for glucose and phenols, but industrial validation remains limited.

A key issue concerns the application of these technologies to non-homogeneous or viscous mash, such as fruit mash, where the physical state and heterogeneity can impair the effectiveness of the sensors.

On the data analytics front, the literature reveals that machine learning is sometimes used in a less than rigorous manner, with small datasets and little attention paid to interpretability: the technologies exist, but they remain fragmented and poorly standardised.

For the sector, the integration of smart sensors and robust data analytics can only lead to greater process control if there is industrial validation and shared standards.

Bibliography: Horváth, T., Kun, S., Sipos, L. et al. Automated monitoring of alcoholic fermentation: trends and challenges. J. Food Sci. Technol. (2026). https://doi.org/10.1007/s13197-025-06528-0