“Real time” quality control of incoming raw materials and conformity assessment of semi-finished and/or finished dietary supplement products.

The main task to accomplish for industrial organizations taking into account the costs and the environmental impact is the expected final product quality. To this aim, an integrated monitoring and control approach, based on the principles of process analytical technology (PAT), can be employed starting from the quality control of incoming raw materials, through the active process control, to final products variability reduction [1]. Near infrared spectroscopy (NIRS), providing non-invasive testing and quick analytical responses, in combination with multivariate data analysis, has shown the potential for performing both untargeted quality control and targeted calibrations of quality attributes [2]. The present study is related to raw materials and was mainly focused on the identity confirmation and/or classification of botanical raw materials, as active constituents in dietary supplement formulations, due to their complex chemical composition. The pattern recognition and classification models developed using principal component analysis (PCA) and partial least squares-discriminant analysis (PLS-DA) showed the potential for further industrial implementation. Moreover, the quantitative prediction of active ingredient(s) in the semi-finished and final products was successfully accomplished by taking advantage of the previously developed linear regression models, obtained using partial least squares (PLS) modelling and the pre-processed NIR spectra. Quick analytical responses regarding incoming raw materials quality as well as real-time assessment of the conformity of semi-finished products can improve the management of resources (natural, human and financial) and allow to plan the manufacturing processes accordingly.