In this paper we present a system for monitoring progress in a mixing and grinding machine via the signal processing of sound emitted by the machine. Our low-cost, low-maintenance system may improve automatic machines and the industrial Internet of Things. We used the Pumpkin Pi board and Raspberry Pi, which are low-cost hardware devices, for recording sounds via a microphone and analyzing the sound signals, respectively. Sound data obtained at regular intervals were compressed. The estimated power spectral density (PSD) values calculated from the sound signals were related to the status of the material during mixing and grinding. We examined the correlation between the PSD obtained by the STFT and the particle distributions in detail. We found that PSD values had both repeatability and a strong relation with the particle distributions that were created by the mixing and grinding machine,although the relation between the PSD and the particle size distributions was not merely linear. We used the PSD values to estimate the progress remotely during the operation of the machine.
ASJC Scopus subject areas