The file incorporates the primary dataset derived from the various fermentations discussed in the article. This includes on-line data acquired in real-time through the control software, which integrates various process variables, as well as offline data obtained from the analysis of samples extracted at different stages of the bioprocess.

The method for estimating discrete rates or time-dependent specific growth rate (µ(t)), specific substrate consumption rate (qS(t)), specific product (or by-product) formation rate (qP(t)), specific oxygen consumption rate (qO2(t)), and specific carbon dioxide production rate (qCO2(t)) was adapted from a study by Cos (2005). Initially, global state variables such as biomass (XV), substrate (SV), and product (PV) concentrations were estimated using off-line data collected during the feeding phase. The Curvefit Toolbox of Matlab R2009a was employed to smooth the curves of these global state variables, and their first time-derivatives were calculated. On-line data on O2 and CO2 concentrations were utilized to calculate the oxygen uptake rate (OUR) and carbon dioxide evolution rate (CER). The specific rates of µ(t), qS(t), and qP(t) were then derived by applying their corresponding mass balances, whereas qO2(t) and qCO2(t) were determined by dividing the calculated OUR and CER by the biomass concentration measured from off-line samples. The uncertainties in these discrete specific rates were quantified through error propagation. As mentioned, the equations for these calculations can be found in the literature, specifically in the work of Cos (2005).