Automation of Microbial Fermentation Process and Detection of Abnormal Behaviors

-
Goal
To develop an autopilot system for microbial fermentation process that optimally adjusts environmental parameters to yield quality microorganisms without human intervention.
-
Our Approach
Based on Machine Learning, MakinaRocks' Intelligent Control solution provides an automation model that imitates operator's past operation by using previous fermenter production data.Furthermore, we build an anomaly detection model that detects sensor abnormality in the fermenter during the fermentation process. In the developmental phase, the initial model is verified and re-learned with the differential data between the control parameters of the machine learning model and the actual control parameters of the operator. After such validation, the fermentation process is executed with a fully automated model.
-
Results
Manpower saving up to 33% compared to the existing operation