Keywords: artificial intelligence, data acquisition, data pipeline, machine learning, oscillations, storage box.

For the storage of small parts and bulk materials, companies usually use storage boxes, which, in addition to the required load capacity, must also be stackable and easily accessible. While a large number of variants are commercially available in the lower price segment, only few approaches exist that enable modern warehouse management. This work describes the equipment of storage boxes with a vibration-based measurement system to detect the filling level. This enables the functionality for intelligent filling level detection as well as for automated reordering of the respective goods. From results of simulations as well as from initial findings of the investigations on a test rig, it was concluded that vibration excitation together with analysis of the natural frequencies is superior to the concept for investigating the decay behaviour. Thus, it was determined that FFT, PSD, and RMS approaches should be pursued. It was found that different crate types, fill levels and filling materials led to a large variance in measurement results and thus to the differentiability of contents and fill level. The respective differences were recognizable in the measurement data, but due to the amount of variation, a manual evaluation was not performed. It was decided that an algorithm based on artificial intelligence should be applied. This work describes the details about the design of a data pipeline in order to process the data. Beside information about the utilized software tools, also details about the artificial intelligence methods as well as the constructive design of the boxes are provided.

download the full-text of this paper (PDF, 6 pages, 527 Kb)

go to the previous paper
go to the next paper
return to the table of contents
return to the volume description