ปัญญาประดิษฐ์

For the first one a typical example is a quality property measured in the lab tests from reel samples, such like tensile strength. You can build a linear statistical model which tells you the dynamics of correlation of different process parameters to the quality property. That enables then further development of algorithms to help you to keep the product quality stable also between the samples and also tells you if the sample is not representative for the entire product.

You can also build a model to describe nominal performance of process, process equipment or an instrument. Once you have created the model to describe the output in normal conditions, you can start continuously monitor the real performance comparing to the model. The devialtions can then trigger alarms which are distributed to applicable systems and persons by SMS messages and e-mails to activate corrective actions.