Method and system that optimizes mean process performance and process robustness

A computer readable medium having a computer readable program embodied in the medium, wherein the computer readable program when executed on a computing device is operable to cause the computing device to:

• accept a definition of one or more independent variables of a process-related experiment;
• receive experimental results observed after the varying of the one or more independent variables during the process-related experiment;
• calculate a prediction model from the experimental results for the one or more independent variables, the prediction model capable of generating predicted results;
• calculate a setpoint for each of the one or more independent variables that optimizes the predicted results to a defined target result;
• define variation around each setpoint under normal operation, the variation comprising a range encompassing each setpoint;
• calculate a process robustness metric using at least the variation(s) around each setpoint, the prediction model, and the defined target result;
• and
• display a representation of the process robustness, at least in part using the process robustness metric.

A computer comprising:

• a processor;
• means for accepting a definition of one or more independent variables of a process-related experiment;
• means for receiving experimental results observed after the varying of the one or more independent variables during the process-related experiment;
• means for calculating a prediction model from the experimental results for the one or more independent variables, the prediction model capable of generating predicted results;
• means for calculating a setpoint for each of the one or more independent variables that optimizes the predicted results to a defined target result;
• means for defining variation around each setpoint under normal operation, the variation comprising a range encompassing each setpoint;
• means for calculating a process robustness metric using at least the variation(s) around each setpoint, the prediction model, and the defined target result;
• and
• means for displaying a representation of a process robustness, at least in part using the process robustness metric.

A high performance liquid chromatography (HPLC) system comprising:

• means for accepting a definition of one or more independent variables of a HPLC-related experiment;
• means for receiving experimental results observed after the varying of the one or more independent variables during the HPLC-related experiment;
• means for calculating a prediction model from the experimental results for the one or more independent variables, the prediction model capable of generating predicted results;
• means for calculating a setpoint for each of the one or more independent variables that optimizes the predicted results to a defined target result;
• means for defining variation around each setpoint under normal operation, the variation comprising a range encompassing each setpoint;
• means for calculating a process robustness metric using at least the variation(s) around each setpoint, the prediction model, and the defined target result;
• and
• means for displaying a representation of a HPLC process robustness, at least in part using the process robustness metric.

A method for evaluating process robustness, comprising the steps of:

• defining one or more independent variables which will be varied in a process-related experiment;
• performing the process-related experiment;
• recording experimental results observed after the varying of the one or more independent variables during the process-related experiment;
• calculating a prediction model from the experimental results for the one or more independent variables, the prediction model capable of generating predicted results;
• calculating a setpoint for each of the one or more independent variables that optimizes the predicted results to a defined target result;
• defining variation around each setpoint under normal operation, the variation comprising a range encompassing each setpoint;
• calculating a process robustness metric using at least the variation(s) around each setpoint, the prediction model, and the defined target result;
• using computer means to display a representation of the process robustness, at least in part using the process robustness metric;
• and
• accepting the setpoint(s) for future processing if the process robustness metric meets or exceeds a threshold, and rejecting the setpoint(s) for future processing if the process robustness metric does not meet or exceed the threshold.