We call a soft sensor a “virtual sensor” whose calculation is based on different real measured variables from real measurement sensors. We design and store a model that is either physically based or generated from data.
For very complex applications, we also use so-called graybox models in which we combine both methods. We also use methods from the field of artificial intelligence to design soft sensors and also use soft sensors in the simulation.
We use soft sensors as support values for the possibility of failure of a real sensor or for measuring points where no real measuring sensor can be used or only with great difficulty.
Innovative measurement technology with soft sensors enables massive cost reductions, for example, in the operation of drying plants or impregnation channels.