Controlling Food recalls through HSI & AI
Updated: May 25
With 2020 behind us, we can look back on the year as a whole and examine the state of food recalls in the food industry. With the increasing number of food recalls, the food industry needs to focus its attention on 2021 and beyond. Though recalls are inevitable, but definitely their frequency of occurrence can be significantly reduced with the help of innovative Imaging and AI technology.
Food recalls in 2020 varied across different types of food products, ranging from fresh produce and perishable foods to packaged foods that were withdrawn from supermarket shelves, warehouses, or distribution centers.
Increasing food recalls is a matter of concern and causing a feeling of fear among consumers at large. Despite the stringent food safety regulation in leading countries, the question arises why these recalls are increasing in spite of rapid innovation in the food sector through technology.
I) What triggers the company to recall the product
Company recalls a product that is not meeting the quality standards and can cause the consumer to become sick. One of the prime reasons behind food recalls is the occurrence of organisms. Aside from the presence of micro-organisms that can cause the consumer to become sick, the presence of a potential allergen beyond a safe limit or mislabeling could also be used as ground for recalling food.
A contaminated product is detrimental to both consumers and companies. Companies risk their brand image if a contaminated product is consumed by the consumers. Consumer trust earned over years can wipe away in a matter of seconds by a recall.
II) FDA Stats
The FDA recalls almost a hundred million units of food every quarter. Prepared foods are the most recalled category, followed by baked goods, vegetables, and beverages. Furthermore, microbiological contamination is the leading reason for food recalls. For instance, salmonella is the most common cause of food-borne illness in the United States. Every year 48 million people fall sick in the US because of food-borne illnesses.
III) Top recall categories
Undeclared Allergen: Top undeclared allergens are wheat, egg, milk, tree nuts, peanuts, and soy
Foreign Materials: Presence of plastic, glass, rubber, bone, wood
Listeria, Salmonella, and E. coli: Recalls due to these organisms constitute a major chunk
IV) Is there a robust solution available that can test the food quality in a few seconds?
Hyperspectral imaging solution is the answer to this. In recent years, hyperspectral imaging has emerged as a promising technology for food safety and quality analysis. This technology was first introduced in satellite and airborne imaging and is now taking its way into the food sector.
Hyperspectral Imaging is a completely non-destructive way to identify different materials or define their properties. Since every material and compound reacts with light differently, the measurements result in individual spectral signatures that can be used to identify materials.
Traditional Digital Imaging is restricted to only 3 bands of light i.e. Red, Green & Blue whereas Hyperspectral imaging is a technique combining spectroscopy and the latest digital imaging technology, where each image is acquired at a narrow band of the electromagnetic spectrum.
As an example, the human eye sees light in three bands (red, green, and blue) of the visible spectrum while hyperspectral imaging divides the spectrum into more bands, typically covering the visible, near-infrared, and short-wave infrared range.
This technology uses an imaging spectrometer to collect spectral information. This device is called a hyperspectral camera which measures hundreds of spectra at each pixel, unlike single-point spectroscopy. The collected spectra are used to form a 3-D image of the target food called a hyperspectral cube.
By analyzing this 3-D image using proprietary Machine Learning algorithms, real-time insights can be gathered on ‘what is in my food, how much and where?'
4 Easy Steps To Detect What Is Not Visible To Human Eye:
Let's explore more about how this is possible using the combination of Hyperspectral imaging and Artificial Intelligence here.