Jul 20th, 2021Digitization and artificial intelligence are finding their way into viticulture
SmartGrape - AI-based MIR measuring system for determining quality in viticulture
With the „SmartGrape“ project, QuoData and the partner institutes aim to analyze the quality determining characteristics of grapes by combining a miniaturized infrared spectrometer with an artificial intelligence (AI) approach. Having an applicability throughout the viticulture process, this reliable quality measurement system will be compatible with both the berries in the wineyard and the mash.
During the project realization, infrared spectroscopy in the mid-infrared region will be used to non-destructively determine the chemical composition and the degree of ripeness of a grape cultivation sample. This system promises to be a powerful analysis method in terms of sensitivity by probing the „fingerprint“ region of many currently interesting molecules.
The measurement system will analyse the obtained complex spectra by employing AI models for reliable characterization of grapes. As the functional relation between parameters defining the quality of grapes are very likely non-linear with multiple interacting variables, machine learning algorithms will be applied. In addition, a targeted approach based on evaluation of marker substances will be employed.
In the „SmartGrape“ project, QuoData will be responsible for developing AI-based evaluation methods to characterize the quality of grapes and mash. QuoData is a recognised expert in the field of statistical quality assurance, especially in the area of food analysis. QuoData has a well-known experience in development of AI models for various spectroscopic and mass spectrometric data in the food sector, for example in the area of food fraud or the identification of bacterial species. With respect to data evaluation, QuoData has regularly addressed aspects of calibration, reproducibility and measurement uncertainty.
In addition to QuoData GmbH, the project also involves the Fraunhofer Institute for Process Engineering and Packaging, the Neustadt Wine Campus at the Dienstleistungszentrum Ländlicher Raum (DLR) Rheinpfalz, Infrared-Process Control GmbH and LiquoSystems GmbH.
The findings of this research project offer new opportunitiers for challenging issues in viticulture, such as authenticity of grapes, influence of vineyard location, effects of climate change, detection of pest infestations and grape diseases, capture effects of climate change by monitoring crop quality over years or check for impact of external factors like soil quality on grape composition and untimately the wine quality.
Launched at the end of April 2021, this project has a scope of 36 months and contributes to the establishment of artificial intelligence (AI) in the fields of agriculture, food chain, and health-conscious nutrition. The „SmartGrape“ project is funded by Bundesministerium für Ernährung und Landwirtschaft.
Further Information: https://www.bmel.de/SharedDocs/Meldungen/DE/Presse/2021/210510-ki-weinbau.html