Skip to Main Content
A sensing method is presented which offers major potential benefits for industrial processes. The method, developed from electrical spectroscopy, aims to discriminate between the component materials in a process and, with calibration data, identify specific materials; for example in batch chemical reactors for the manufacture of pharmaceutical products. The method addresses the key design need to complete a sensing operation within a temporal window that allows for the process dynamics. Two key system requirements are considered. First, a review is included of candidate excitation signals, in terms of the major parameters of interest, for the relevant frequency range. Second, the acquisition of the corresponding response signals and the extraction of the spectroscopic data from which the materials of interest may be identified. This is based upon an algorithm which is introduced based on the wavelet transform. The composite method is illustrated in trials using a process impedance simulation model and experimental tests on a crystallization process. This paper offers conclusions for applications of the fast sensing method to characterize different process (and other) materials.