Agriculture, from ENEA system for selecting ‘good’ hazelnuts with Terahertz radiation and AI

Industry impact on local water resources new methodology from ENEA

(Finance) – AENEAS has developed technology capable of quickly identifying hazelnuts faulty without shelling them, in order to eliminate them before they enter the industrial processing chain. The system, composed of a source Terahertz (THz) in solid state and by a detector, it is able to measure the water content in hazelnuts, a qualitative parameter considered very important for the conservation of the product by companies operating in the sector. The efficiency of this technology has been successfully verified in the laboratory, but thanks to its simplicity, it can be easily implemented in an industrial system applicable to the hazelnut supply chain. Furthermore, the technology has been further perfected with a system of artificial intelligence (AI), implementing a neural network on a large number of samples.

“The cultivation and processing of hazelnuts is one of the excellences of the Italian agri-food sector, which sees companies of the most varied sizes involved in its supply chain, from the small farmer to the large multinational”, underlines Manuel Greek, who is carrying out his PhD on the topic in a collaboration between ENEA and the University of Roma Tre. “The sector – she adds – requires finding increasingly efficient methods to identify spoiled fruit, which when put into production can ruin the quality of entire batches of hazelnuts. At the moment, the selection procedure can only be automated for shelled fruits, otherwise it is only available via non-automated visual analysis.”

Already used for applications in field nuclear and of conservation of cultural heritage, the system developed in the ENEA laboratories in Frascati is based on the characteristics of THz radiation, which has frequencies slightly higher than those used in common microwave ovens. “Electromagnetic radiation, at these frequencies, is able to easily pass through dielectric materials, such as the hazelnut shell, and detect the amount of water inside the fruit,” he explains Emilio Juvenal, ENEA researcher at the Plasma Applications and Interdisciplinary Experiments Laboratory. “By measuring the transparency of a series of hazelnuts to THz radiation, it was possible to verify that the damaged ones were much more transparent than the healthy ones and this even when from an external visual analysis the fruit appeared perfectly healthy,” adds Giovenale. “This system improves the recognition and allows us to better define the threshold for discrimination between healthy and damaged samples, on the basis of requirements supplied by those who will then have to use the hazelnuts”, he concludes.

In addition to applications to identify failed fruit, the technology can be used to monitor the complete fruit growth cycleuntil maturation, by measuring the water content, thus obtaining useful information for its optimal conservation after harvesting.

tlb-finance