| CPC G05B 23/024 (2013.01) [G05B 23/0221 (2013.01); G06F 18/214 (2023.01); G06N 3/04 (2013.01); G06N 3/08 (2013.01)] | 10 Claims |

|
1. A method to provide real-time quality control of a culture for bioproduction, wherein the culture takes place in one of: a bioreactor, and a fermenter, comprising:
receiving real-time process data for one or more of a set of parameters for a current production run;
based at least in part on the received real-time process data, generating a prediction using an instance of a machine learning model, wherein the machine learning model is a neural network, the instance being uniquely associated to a particular culture day and having been trained on process data from past production or development runs;
rendering a visualization, the visualization having a timeline of the current production run, wherein based on the prediction the visualization depicts at least one of:
an early culture failure detection point, and a mature culture endpoint detection and feature optimization detection point, wherein each such detection point occurs on the timeline prior to an anticipated last day of the current production run;
based on the prediction, generating signaling to control advanced process control (APC) tooling associated with the bioreactor or fermenter; and
applying the signaling to the APC tooling, thereby initiating an interdiction in the bioreactor or fermenter to early terminate the current production run.
|