Wine fermentation is unique from other fermentation industries as the initial conditions of grape juice
chemistries vary drastically from year to year and vine to vine, resulting in widely different fermentation
outcomes. Despite an entire industry of new yeast strains and yeast nutrients, incomplete or slow
fermentations still exist in commercial winemaking. Incomplete and slow stuck fermentations are costly
and slow down winery operations. In previous work, we have demonstrated a differential pressure
technique to measure density, a parameter that allows wine fermentation to be monitored. In this
work, we extend the automated density measurements to a model that describes the kinetics of wine
fermentation. Using numerical methods, the best parameters in the model (specific maintenance rate
of yeast, lag time, initial nitrogen, viability constant and ethanol inhibition constant) are found that
minimize the sum squared residual between the measurements and model. Combined with automated
density measurements, an on-line model estimation system would allow, across all concurrent
fermentations in a winery, the automated prediction of fermentation density and the quantification of
rates of energy and CO2 evolution. In this work, a software architecture is presented to record sensor
data, estimate the model parameters in real-time and visualize the results. A new numerical method
technique based on direct grid search was evaluated on a set of red and white fermentations. From the
rate of fermentation, the rates of energy and CO2 production were also compared to the on-line
modeling predictions. As more data becomes available, the confidence in the on-line predictions
increases and closely follows the automated measurements. Finally, the summation of overlapping
energy and CO2 loads were calculated from automated density measurements for all fermentations
from the 2021 harvest at the UC Davis Teaching and Research Winery. This work was supported by the Rodgers University Fellowship in Electrical and Computer Engineering (JN) and Stephen Sinclair Scott Endowment in Viticulture and Enology (RB).