Pocket scanner blasts food with infrared light to determine its freshness
Every year, millions of tonnes of perfectly good food are wasted around the world because people aren't sure it's still fresh. To cut down on this, researchers at Fraunhofer are developing an infrared pocket scanner that will let consumers, supermarkets and other food handlers determine if a food item has gone bad and even its degree of ripeness.
According to the United Nations Food and Agricultural Association, one third of all the food produced in the world is lost or wasted. That's about 1.3 billion tonnes annually. In the developed world, which has the highest percentage of waste, that works out to US$680 billion dollars down the garbage disposal each year.
There are a number of reasons for this wastage, but a large factor is that consumers tend to have trouble determining if a food item is still edible. Often meat and produce will be judged on aesthetic grounds, or the "sell-by date" will be confused with a "use-by date."
To combat this, the Bavarian Ministry of Food's "We Rescue Food" alliance, has launched 17 initiatives, one of which is the Fraunhofer pocket scanner. The device is designed to be an inexpensive solution for determining the edibility and shelf life of foods from farm to table.
The scanner is based on a high-precision near-infrared (NIR) sensor. An infrared beam is shined on the food and the reflected light is measured across the IR spectrum. By comparing the absorption spectrum from the food with that of a known sample, the device can determine not only if the food is still edible, but also its ripeness and even if it's a counterfeit, such as trout being passed off as salmon.
It's a technique already used in laboratories, but the tricky bit is reducing the size and cost of the device without sacrificing function. This is managed by using new, small, inexpensive sensors.
According to Fraunhofer, the scanner is still in the demonstrator stage. Currently, it can only handle homogeneous foods, so it can analyze a potato, but not a pizza with its many toppings. The hope is that hyper-spectral imaging and fusion-based approaches using color images, spectral sensors and other high-spatial-resolution technologies may overcome this in the future.
Another aspect under development is a machine-learning algorithm that will allow for better pattern recognition. So far, the team has worked with tomatoes and ground beef using statistical techniques to match the NIR spectra with the rate of microbial spoilage and other chemical parameters, allowing them to measure the germ count and the shelf life of the meat.
This is done by sending the scan data over Bluetooth to a cloud database for evaluation. The results are then sent to a mobile device app to show if the item is still good, how much shelf life is left, and tips on how to use the food if its sell-by date has expired.
The research team says that supermarket tests are slated for later this year to see what consumers think of the scanner. The technology can not only be used for foods, but also for wider applications, such as sorting plastics, wool, textiles, and minerals.