Fruit and vegetable scanner for wholesale quality assessment

Precisely record quality - waste less food

Around one third of the food produced worldwide ends up in waste. A large proportion of food waste is generated in the food industry. One of the reasons for this is that wholesalers have difficulty predicting the shelf life of fruit and vegetables. As a result, fruit and vegetable wholesalers often have to throw away some of their produce, reducing their profits and causing unnecessary food waste.

Our multisensory fruit and vegetable scanner offers potential for comprehensive and precise quality detection and freshness prediction directly at goods receipt.

Determining the degree of freshness and ripeness of fruit and vegetables with the help of various sensors and artificial intelligence opens up new possibilities for the forward-looking handling of food. Among other things, the shelf life of food can be better predicted. Improved control of the time of sale and pricing for various goods enables retailers to act more efficiently. There is also the possibility of sorting out low-quality goods, which can extend the shelf life of the remaining products. This should make it easier for retailers to optimize the flow of goods and avoid food waste. In this way, the scanner helps to reduce costs while improving the ecological impact in the food industry.

Automated quality assessment using AI

In wholesale, screening of incoming fruits and vegetables is essential for continuous quality assurance. The scanner is designed to automate, standardize, specify and expand this quality control. The sensor data is evaluated with the help of artificial intelligence. The system learns from previously collected data and can thus adapt flexibly to different problems with the help of machine learning.

The system: multimodal sensor technology, local evaluation

The multimodal sensor system is currently equipped with microwave sensors (radar), a balance, and a five-channel VIS-NIR area scan camera with two wide near-infrared bands. A broadband LED light source is used for illumination. The system can be supplemented with additional or other sensing equipment. The fruits and vegetables can remain in the crates for measurement, which are detected with the push of a button. An additional barcode scanner allows linking with information from the delivery bill. The touch interface allows easy operation of the system. The evaluation takes place locally on the device through machine learning.

Our goal

Our goal is to continuously improve the scanner prototype in order to optimally align it with the specific requirements and fields of application in fruit and vegetable wholesale. We strive to generate maximum benefit and added value for retailers with our research work on the one hand, and to counteract food waste at the same time on the other. We are always open to collaborations and research opportunities that support our vision.

Research partners

• Fraunhofer Food
• Fraunhofer CCIT
• Fraunhofer IZFP
• Fraunhofer IVV