Soft-sensor development for monitoring the lysine fermentation process.

Monitoring cell growth and target production in working fermentors is important for stabilizing high level production. In this study, we developed a novel soft sensor for estimating the concentration of a target product (lysine), substrate (sucrose), and bacterial cell in commercially working fermentors using machine learning combined with available on-line process data. The lysine concentration was accurately estimated in both linear and nonlinear models; however, the nonlinear models were also suitable for estimating the concentrations of sucrose and bacterial cells. Data enhancement by time interpolation improved the model prediction accuracy and eliminated unnecessary fluctuations. Furthermore, the soft sensor developed based on the dataset of the same process parameters in multiple fermentor tanks successfully estimated the fermentation behavior of each tank. Machine learning-based soft sensors may represent a novel monitoring system for digital transformation in the field of biotechnological processes.

Data science-based modeling of the lysine fermentation process.

Mathematical modeling of the fermentation process is useful for understanding the influence of operating parameters on target production and control performance, depending on the situation, to stabilize the target production at a high-level. However, the previous approaches using physical modeling methods and traditional knowledge-based methods are difficult to apply on working fermentors at a commercial plant scale because they have unknown and unmeasured parameters involved in target production. This study focused on developing an ensemble learning model that can predict the amino acid fermentation process behavior based on observation values, which can be obtained from fermentation tanks and future control input. The results revealed the influence of each control input on lysine production during the culturing period. Furthermore, high-order stability, which achieved the target trajectory for lysine production, was realized using dynamic fermentation controls. Additionally, this study demonstrates that the fermentation behavior on a commercial plant scale is reproduced using the ensemble device. The ensemble learning model will provide novel control system with data-science based model of Industry 4.0 in the field of biotechnological processes.