The PicknPack project aims to provide the food industry with new perspectives for a more flexible, more automated, gentle and robot-based production where smart factories are concerned. A PicknPack system consists of different modules, which can be configured as desired, or extended by additional units depending on the order and customer requirements. The QAS module (QAS, Quality Assessment and Sensing) here plays a central role. To see it work, visit the VDMA Technology Lounge in hall 5 / J38, where the VDMA presents ideas and solutions for tomorrow’s machines in cooperation with partners from industry, science and research.
Quality assurance is one of the most important prerequisites for guaranteeing customer satisfaction. The universities of Leuven (Belgium), Wageningen (the Netherlands) and Manchester (United Kingdom) as well as the Danish company InnoSpexion ApS all were involved in the development of the QAS module. It contains several submodules for different test and control procedures. This includes five scanning methods: an X-ray scanner, a hyperspectral imaging system, an RGB line scanner, a 3D imaging system and a microwave sensor unit. Based on the experience gained from previous projects, the developers have created a food database that allows the QAS module to be easily trained to recognize quality features of different products. The demonstrator shown in the VDMA Technology Lounge demonstrates working with the methods of hyperspectral and 3D imaging.
The Hyperspectral Imaging Sensor used here detects wavelengths from 600 nm to 1000 nm and covers spectral bands in the visible red and near infrared (NIR) regions. The information obtained from the image analysis makes it possible, for example, to evaluate the degree of maturity as well as surface defects of tomato panicles or grape vines or other fruit and vegetables. The advantage of the Hyperspectral Imaging method is the excellent detection rate of surface defects compared to conventional RGB cameras or the human eye.
The 3D Imaging System, on the other hand, operates with two 45° inclined sensors based on the principle of laser triangulation, which capture 50 percent of the product surface (the upper half). The data obtained in this way can be used, for example, for the exact calculation of the size and weight of the scanned products. Another use of this data is controlling the pick robot. On the basis of the recorded data, the robot knows the exact position, size and orientation of the products on the conveyor belt or in the transport boxes. This allows it to align its gripper perfectly, to take the products gently and place them in trays.
Interested visitors will be able to find out about these and other innovations as well as practical research results from the partner companies at the VDMA booth in Hall 5 / J38.