Abstract
The criteria for application of models in industry are based on the relevance of the results obtained for the needs of the project. “Fitness for purpose” is key. A model should be at the appropriate level for its intended use, neither too simplistic to give reliable results, nor at an unnecessary level of detail that requires irrelevant input data and extends calculations. A distinction is made between models that can be used regularly in process operations – often relatively simple ones – and those that can be used once to give in-depth process understanding. Examples of industrial application of appropriate models are given, particularly in pharmaceuticals. One major challenge is to verify model performance and demonstrate that it is reliable and robust. This can be achieved in several ways, including comparison with data-driven statistical software.
Disclosure statement
No potential conflict of interest was reported by the authors.