As a stakeholder in the food supply chain you need to ensure the quality and authenticity of your products. Therefore you have elaborated an extensive quality management system that takes into account all aspects and responsibilities. However, evidence shows that the number of quality and fraud incidents in the food industry are enormous. The figure below shows that the number of reported incidents in the EU RASFF portal is increasing and reaches more than 2000 incidents/year!
Reference: Joseph Timkovsky: https://www.linkedin.com/posts/joseph-timkovsky_food-foodfraud-foodintegrity-activity-6588002506410860544-1HOW
The reported events are due to mislabelling, administrative shortcoming or based on wet lab analysis of food samples. It is generally thought that this only represents the tip of the iceberg. That’s because only limited samples are measured. The amount of internal measurements for example at retailer side is very limited compared to the amount of product that is handled. They even do not control food fraud by means of chemical analysis on very known examples of fraud in spices, olive oil, meat, fish. They often assume everything is fine if the paperwork is fine.
Chemical analyses are very expensive and take much time when they are to be deployed on a large scale. Therefore Xpectrum uses NIR spectroscopy. What does this offer?
Hands-on technology which is very easy to use and suitable for your operators at reception, in warehouses or in production
Immediate result: thanks to machine learning and AI methodologies the operator will receive immediate result. Incomparable speed with chemical analysis!
Non-destructive: samples do not need any treatment, they can be measured as such
Scalable: measuring more does not cost you more! The more you scan, the more you know, all at the same cost! That's what we want you to do!
A 6-month pilot test at a retailer has shown this competitive advantage of NIR spectroscopy. Two products were selected. During this period the retailer would have sent around 8 random samples to the lab for chemical or DNA analysis, depending on the product. With our cloud solution they measured everyday with a total of 1600 samples! Our AI-powered calibration model, operated in the cloud, detected 20 suspicious samples (around 1%). To validate the result we have sent these samples for wet lab analysis. And guess what, 80% of these were confirmed to be out of spec. With this system the retailer was able to increase its monitoring scheme 200-fold (!). Instead of selecting random samples to be sent to the lab, the system told him which samples he should investigate further, which gave him a 80% chance of hitting the target. The difference in analysis cost was estimated to be €30k for each product if all these 1600 samples had to be sent to the lab.