Subscribe to our Newsletter
The LAA is accredited by the National Accreditation Entity (ENAC) in Spain, according to UNE-EN ISO/IEC 17025 Standard, to carry out physical-chemical, sensory, microbiological and biomolecular tests of agri-food products and acts as the control inspection body in Spain. It is located in Madrid (Spain), has more than 80 employees, and occupies an area of 7,000 m2 distributed in two buildings, of which 85% is intended for laboratories.
The LAA is attached to the General Subdirectorate of Control of the Food Quality and Agri-Food Laboratories, within the General Directorate of the Food of the General Secretariat of Agrarian Resources and Food Security.
Honey quality and authenticity assessment with pollen analysis is done manually with a microscope. There is no expert equipment to carry it out as with other parameters such as pH, moisture, chemical analysis, or sugar composition. The identification (species classification) and counting of pollen species with direct visual inspection with a light microscopy on the honey’s sediment still is a traditional technique that proves technically very complex, time-consuming, exhaustive, cost-intensive and prone to human error. Within a unique analysis, it is required to count and classify, around 500 pollen grains. The Challenger sometimes cannot afford high demands and workload required in LAA because they also have to analyze many other foods, not only honey.
At the Agri-Food Arbitration Laboratory (LAA) their expert team of technicians must carry out tenths of honey analysis from different botanical origins and geographical regions and this task implies exhausting work and high level of expertise. The Spanish Ministry aims to increase the control inspections for honey at national level, with the aim to reduce the incidence of fraud in Spain and better target the current problem of honey’s correct labelling. However, the workload related with honey analysis, combined with the difficulty and complexity of doing such challenging tasks, seriously complicate the objective a lot.
Microfy Systems is a company located in Barcelona, Spain, which develops, manufactures and commercializes affordable AI-powered automated microscopes for the agrifood industry for quality control procedures. A multidisciplinary 8-people team of engineers and technicians are behind each commercialized invention, and currently cover all the manufacturing steps within the supply chain, from the concept to commissioning, on a B2C approach. Our team has 400m2 floor space that includes a technical office, a mechanical workshop, an electronics lab, a production area and prototypes room plus multiple workplaces, and a testing area.
Our past involvement in the honey industry has led us to identify one necessity, the need to automatize quality analysis of honey, pollen, honeydews. From this need, Honey.AI was born [www.honey-ai.com], offering an automated solution specially designed for cooperatives, and traders, and packers who want to benefit from AI and innovate in a traditional sector.
Honey.AI is an automated digital microscope that allows autonomous pollen analysis in honey samples thanks to a robotized sample scanning step coupled with an image processing pipeline in the cloud with widely trained deep learning models. Honey.AI automatically scans the sample and uploads thousands of pictures to the cloud.
The AI module counts with a database of more than 100 pollen species already trained, more than 200,000 images, and allows the user to analyse honey samples without any human aid.
With this project it will be assessed how to use HoneyAI for screening of fraud in Spanish honeys (botanical and geographical).
Honey.AI was conceived as a response to automate the tedious and complex work of pollen analysis using computer vision and neural networks within an affordable robotized microscope. Initially, the device was intended to be sold to honey packers, beekeeping cooperatives, traders and apiculture associations to perform quality analysis on a B2C business model. Nowadays, those honey stakeholders spend thousands of euros per year to conduct this kind of analysis in specialized large laboratories and have to wait 1-2 weeks to receive the results back, so they lose control of the product and the process.
Honey.AI allows on-site measurements, offer affordable prices at 12€/test, immensely reduces human dependency and human error, and provides results almost real-time. The device also allows the user to conduct other quality analysis, such as colour grade, yeast concentration counting, honeydew analysis, starch detection, etc.. which also can be complementary data for the end-users
This project has received funding from the European Union’s Horizon Europe research and innovation programme under the Grant Agreement 101071212.
Views and opinions expressed are however those of the author(s) only and do not necessarily reflect those of the European Union or European Innovation Council and SMEs Executive Agency (EISMEA). Neither the European Union nor the granting authority can be held responsible for them.
