(Predictive) model development
If you don't already have your own predictive model ("soft sensor") or supervisory control sheet, our APC engineers and data scientists can develop one for your application.
What Is a soft sensor ?
Soft sensors (also called virtual sensors or inferential models) are widespread in the process industry and are used for on-line prediction, monitoring and control.
They use a mathematical computer model to continuously derive the values of hard-to-measure product or process properties from easy-to-measure variables such as temperature, pressure, humidity, flow etc..
Various soft sensors are in use today in industry to estimate difficult/slow/expensive-to-measure variables such as polymerisation rate, fermentation progress, concentration, moisture or other properties.
A typical example is the use of soft sensors to have every few seconds a prediction of a property that otherwise can only be measured in the lab after long processing on a sample that is taken from production.