Greater efficiency and safety in food production : Date:
When it comes to food, the quality has to be right. If the apple juice is too sweet or too sour, production is stopped. Any deviation must be detected immediately. To this end, researchers in the QSFood project are developing digital models – with funding from the Federal Ministry of Education and Research.
How fresh apple juice must taste, smell and look, is precisely defined in the food production industry. Experts use a sensory analysis to check properties such as colour, odour, taste and ripeness. In addition, technical chemical analyses are carried out. There are also legal requirements, such as maximum quantities of microorganisms.
If the quality controllers find grounds for rejecting the products, often a whole batch has to be destroyed, which means the loss of raw materials, energy and time. This is where a research team at Ostwestfalen-Lippe University of Applied Sciences and Arts in Lemgo comes in: the idea is to use data related to raw material and sensory measurement in the production process to detect deviations from quality standards at an early stage. Data from various sources are aggregated on a big-data platform. The aim of the project is to develop a digital model of the intermediate products in the production chain, using self-learning algorithms to determine their quality. This should then alert producers to deviations even before the food tests.
The researchers see an opportunity to use special algorithms to extract more detailed information from measurement data than has previously been possible. For example, measurements in a near-infrared spectrum can be used to derive the sugar content of a beverage, possibly even the acidity, turbidity or other quality indicators as well. The researchers plan to collect these data in addition to the prescribed laboratory analysis. In doing so, it is also evaluated whether an added value can be gained from the data of the already installed sensor technology. Using digital models of intermediate products, it should soon be possible to detect defective batches in real time. The economic advantages are obvious: the raw material yield can be increased by adjusting the processes in time. Safety risks, such as foreign substances in muesli, can also be immediately detected and eliminated. This means that the research results are of particular benefit to small and medium-sized enterprises, which are less able to convert their production facilities than large firms.
The Federal Ministry of Education and Research is supporting the QSFood research project with more than one million euros within the framework of the FH-Impuls funding line as part of the Research at Universities of Applied Sciences programme.