Currently, producers and retailers of perishable products face the challenge of producing fresh products that are safe and of high quality. Product quality of such products is affected by many factors, including but not limited to microbial activity and oxidation effects. Due to the lack of fast, non-invasive methods for measuring product quality, it is challenging for producers to assign custom shelf-life of their products based on the quality of input materials and/or the quality at the time of packaging. Therefore, a relatively large ‘margin of safety’ is necessary, which limits the producers in flexibility in planning of production, cleaning, storage and transport to the retailers. For the retailers, inaccurate shelf-life designations limit the possibilities for selling these products, which results in unnecessary food waste and negative impact on societal, environment and economic level.

Spectral imaging (SI) is a powerful analytical technique that captures images across multiple wavelengths of light, allowing for fast and detailed identification and characterization of materials based on their unique spectral signatures. This makes SI a very attractive sensing technology for a wide range of applications. In addition, SI offers the potential to measure through transparent packaging, making it a potentially non-invasive methodology for measurements. Spectral imaging is already being explored as technology for non-invasive quality assessment of fresh fruits, where product deviations can be detected before blemishes on the outside can be observed.
Within SPECTRA, the aim is to explore the possibilities of SI further in a focussing on microbiological quality of pork meat. Besides microbiological data, several other quality parameters will be included (e.g. colour, pH) that are routinely measured as product parameters (by the private partner). With the use of artificial intelligence, all information will be linked with the goal to establish a predictive tool for pork meat quality based on spectral imaging output. The developed predictive model for pork meat will serve as a demonstrator for the applicability of spectral imaging as predictive read-out for shelf-life and give an outlook for opportunities for other perishable products. Having a validated SI method in place would be a large benefit to producers in helping them to optimize planning, production process, logistics in the supply chain. If the pork meat case study explored in the SPECTRA project is successful, the technology can be adopted for other perishable food products such as fish or other fresh meat or meat analogue products.

The aim of this project is to develop existing spectral imaging (SI) technology in UV, visible, NIR range combined with
fluorescence for application into rapid and non-invasive analysis of microbiological product quality of pork meat. SI has a
major advantage in detection of microbiological quality in terms of speed and non-destructive nature in comparison to
current, cultivation-based methodology.

The SPECTRA project fits with KIA mission 4. “Sustainable valued food, which is healthy, accessible and safe”, and specific to
innovation program 4B. “Sustainable processing and food safety, fresh and processed”, specifically subprogram 3.
“Improvement of sustainability of food processing by reduction of FLW”. The project will contribute to food processing
becoming more sustainable by using food resources more efficiently and wasting less. It will develop relevant technology
that can be applied at scale in operations, ensuring quality and safety throughout the supply chain.

The expected results for this project are:
▪ Insight in the ability of spectral imaging to assess (microbiological) product quality show-cased for pork meat
▪ Validated use of spectral imaging for an industrial use case
▪ A draft predictive model for product shelf-life of pork meat based on spectral imaging