Application of mechanistic modelling and machine learning for cream cheese fermentation pH prediction

Abstract

BACKGROUND: Cheese production occupies a small but growing share of the current dairy industry. Industrial cream cheese production involves the complex scheduling of multiple batch fermenters and various downstream units. However, significant end-time variation from different fermentation batches makes downstream scheduling challenging and thus decreases the process throughput. RESULTS: This research addressed this challenge by using an artificial neural network (a Long-Short Term Memory Network, LSTM) in combination with a mechanistic model describing the changes in biomass, lactose, and lactic acid concentrations. The LSTM network/mechanistic modelling approach shows an end-time difference of 3 min over batch times of 6 to 7 h compared to the laboratory experiment, with an overall accuracy (R 2) of over 0.99. CONCLUSION: The proposed hybrid fundamental/artificial neural network (ANN) model framework could reasonably predict the cheese fermentation pH with limited data. The outcome of this research enables fermentation end-time prediction, makes the downstream scheduling possible, and thus, could help to improve the process throughput. The developed model could also be used for further cheese digital twin development, based on how better product quality control and higher process efficiency could be achieved.

Robustness: linking strain design to viable bioprocesses

“…Starter seed cultures are stored in freezers, a condition that might cause cell viability and vitality to vary between batches and vials [72]. Therefore, both the dry weight and activity of cells in the initial bioprocess inoculum should be considered [73]. Viral, bacteriophage, and bacterial contamination can severely impact on both mammalian and microbial bioproduction (e.g., leading to premature shutdown of bioprocesses); this can be avoided through rigorous PCR-based testing of raw materials [74], the use of antibiotics [75], or by engineering strains which efficiently assimilate xenobiotic media compounds and thus outcompete contaminants [76].…”

“…Large datasets can be used as basis for emerging artificial intelligence to predict and control bioprocesses. In cream cheese production, an artificial neuronal network together with a mechanistic model have been developed to predict overall fermentation time using only the initial biomass, lactose, and lactic acid concentrations as variables [73]. A hybrid framework integrating data-driven methods with a genome-scale metabolic model demonstrated promising accuracy when predicting E. coli performance under typical bioprocess conditions and available pathways [91].…”

Machine Learning in Chemical Product Engineering: The State of the Art and a Guide for Newcomers

“…Accordingly, hybrid modeling approaches, combining both knowledge-based and datadriven characteristics display increasing popularity in solving problems where the mechanisms are too complex to be exhaustively described mathematically or where the relevant knowledge/understanding of the phenomena prevailing on a specific part, or range of con-ditions, of the process is missing. Numerous relevant applications have been reported in the food industry [54][55][56], biopharmaceutical industry [57,58], cosmetic products design [59], catalysts design and discovery [21], reaction prediction [60] and polymer processes [61][62][63][64].…”

Which model? Comparing fermentation kinetic expressions for cream cheese production

“…[ 8,19 ] Li's GBM feeds back the corresponding predicted concentration and pH value by inputting the initial biomass, lactose, and lactic acid concentration during the fermentation in cream cheese production. [ 20 ] In another GBM, the initial fermentation condition (lactose concentration) and operational conditions (temperature, pH, reactor stirring rate, and reaction time) were used as inputs to an artificial neural network (ANN) to optimize the parameters. Those parameters were then applied to calculate the current concentration (biomass, lactose, lactic acid) based on mass balance.…”