Automated non-invasive Detection of Surface Contamination and Optimizing Surface Biofilm Removal Conditions in Large-Scale Food Processing Equipment

Approach:
Ultra-rapid detection of food contamination is important to effectively use mitigation strategies and minimize foodborne illnesses. Scanning-based image spectrum instantly detects biological contamination such as fecal matter or bacterial biofilms. Fecal matter may potentially indicate contamination with highly pathogenic avian influenza (HPAI) on shelled eggs. The research will extract an innovative spatial relationship of pixels -based image texture features and spectral wavelengths of interest to build a joint optimization function. The neural network classifier will be developed to detect biofilm on biotic and abiotic surfaces. Multi-species biofilms, where foodborne pathogens form biofilm on surfaces with naturally occurring bacteria in food processing environments, are extremely difficult to remove by commonly employed sanitizers. These pathogens contaminate food during subsequent processing which may result in foodborne illnesses and/or expensive food recall. A non-invasive particle image velocimetry (PIV) system will be used to characterize fluid-flow dynamics of sanitizer under hydrodynamic conditions. Predictive model will be developed for efficient removal of resistant biofilms from equipment surfaces.