The Catalan company Sonicat Systems, based in L'Hospitalet de Llobregat, has developed a digital microscope with artificial intelligence that analyzes honey and automatically determines its botanical origin and assesses its quality. The technology is based on a world-pioneering software that, using an internal database of more than 180,000 images, quantifies the pollen profile of honey with virtually no human intervention, thus ensuring correct labeling.
Sonicat Systems has received support from Acció, through the Cupons d'Indústria 4.0 program, grants aimed at implementing new technologies in industrial processes, explained the Department of Business of the Generalitat.
"According to sector regulations, for a honey to be labeled as monofloral, it must generally contain certain minimum concentrations of pollen from that botanical species," recalls the co-founder and CEO of the company, Irache Perales. "With the system developed by Sonicat Systems, called Honey.AI, the goal is for the beekeeper, the packaging company, or the cooperative that produces the honey to be able to analyze the product without needing to send it to an external laboratory, thus saving time and money," explains Perales, reports ACN.
The system developed by the Catalan company is based on a simple microscope that they have robotized, to which they have also incorporated a camera, which is controlled by a desktop application. "We have trained an artificial intelligence model from a database with more than 180,000 photographs of different types of honey from all over Europe and various pollen species so that the system learns to classify them", says the CEO of Sonicat Systems. The user only has to insert a slide with a small sample of centrifuged honey into the microscope, which automatically analyzes it and takes images that it then processes and classifies in the cloud using the software created by the company.
The system will soon also be able to have an automated dispenser that introduces up to 24 honey samples into the microscope, which can be analyzed autonomously throughout a day. In this way, the botanical species from which the honey has been extracted is determined to be able to certify the quality of the product automatically. "There is no other system like it in the world," remarks Perales. "Universities have promoted research to automate images and there are programs to analyze pollen allergens in the air, but these technologies had never been applied to honey analysis in the commercial field, and with an automated device," points out the executive director of Sonicat Systems.
First design with Eurecat and commercialization this 2023
The company has collaborated in recent months with the Eurecat technology center to develop the first design of the automatic dispenser Honey.AI and, currently, is carrying out the first pilot tests of its microscope with key companies in the sector, such as Granja San Francisco or Apisol. They already have a functional prototype and are finalizing the process to obtain CE marking. Sonicat Systems plans to start commercializing its product during the first half of this 2023 and begin international expansion from next year, with the markets of France, Portugal, Italy and Germany as first targets.