Different conveyor belt materials used by the meat and other food industries were compared, regarding their cleanability as bacterial reduction rates in relation to their surface topography. Eleven thermoplastic polymers, four stainless steels, and five aluminized nanostructured surfaces were investigated under laboratory conditions. Cleanings were conducted with water only, and with an alkaline foam detergent. Overall, scanning electron microscopy revealed remarkable differences in the surface topography of the tested surfaces. Water cleaning results showed that nanostructured aluminized surfaces achieved significantly higher cleanability rates compared to the eight thermoplastic surfaces, as well as the glass-bead blasted rough stainless steel. Thermoplastic surfaces showed overall low cleanability rates when cleaned with alkaline detergent, while stainless steel and nanoporous aluminum showed high variations. Overall, nanoporous aluminum showed promising results as it can be used to coat conveyor belts. However, compatibility with cleaning detergent and sensitivity to scratches must be further investigated. Overall, it can be concluded that cleanability is not only influenced by surface roughness, but also by the overall surface finish, scratches, and defects.

The high perishability of fresh meat results in short sales and consumption periods, which can lead to high amounts of food waste, especially when a fixed best-before date is stated. Thus, the aim of this study was the development of a real-time dynamic shelf-life criterion (DSLC) for fresh pork filets based on a multi-model approach combining predictive microbiology and sensory modeling. Therefore, 647 samples of ma-packed pork loin were investigated in isothermal and non-isothermal storage trials. For the identification of the most suitable spoilage predictors, typical meat quality parameters (pH-value, color, texture, and sensory characteristics) as well as microbial contamination (total viable count, Pseudomonas spp., lactic acid bacteria, Brochothrix thermosphacta, Enterobacteriaceae) were analyzed at specific investigation points. Dynamic modeling was conducted using a combination of the modified Gompertz model (microbial data) or a linear approach (sensory data) and the Arrhenius model. Based on these models, a four-point scale grading system for the DSLC was developed to predict the product status and shelf-life as a function of temperature data in the supply chain. The applicability of the DSLC was validated in a pilot study under real chain conditions and showed an accurate real-time prediction of the product status.