Catalogue of Services

FEM will support companies developing machine vision tools (both hardware and software), e.g., for counting and estimating the size of fruits, either by collecting ground-truth data or by comparing the results of the FEM algorithms with the customer solution. These activities are supported by both physical and digital facilities.

AI model training

Testing of proximal mobile field sensing technology

Field testing of proximal sensing technologies onboard mobile systems (i.e., UAVs and robots) for the assessment of plant physiology traits. FEM will support the assessment and validation of proximal sensing solutions by hosting field trials at the physical facilities available within the FEM Campus (i.e., smart vineyard, micro-vinification cellar, grapevine collection, and fish farm) and supporting the activities by using equipment to collect ground-truth data. As an example, FEM has developed and has available in the field tools used to identify water stress situations, based on the measurement of canopy temperature.

AI model training

Testing of remote sensing technology

Field testing of remote sensing technologies (i.e., MODIS, Sentinel data, PRISMA) for the assessment of plant physiology traits (canopy temperature, vigour) or diseases. FEM will provide time series of NASA-MODIS and ESA-Copernicus satellite data, experimental fields, and instruments (i.e., spectroradiometers Spectral Evolution RS-5400, ASD Fieldspec PRO FR, etc.) to collect ground-truth data for the validation of remote-sensing models. In particular, FEM has 20 years of daily MODIS Land Surface Temperature images and weekly MODIS NDVI-EVI vegetation index maps. FEM has also been acquiring Sentinel-2 multispectral data at 10 m spatial resolution since 2017.

AI model training

Testing of sensor technology for irrigation scheduling

Field testing of IoT prototypes to monitor soil features and comparison to research-grade sensors FEM will support companies developing IoT solutions, e.g., irrigation scheduling solutions, either by collecting ground-truth data or by comparing the results of the algorithm with their own models. The areas in which FEM is active concern irrigation scheduling, frost alerts, phenology, crop water stress identification, etc. These activities will be supported by the FEM Campus facilities, research-grade weather station data, high-performance computing machines, and LoRaWAN technology for data transmission.

Testing of sensor technology in the wine-making process

Location

Italy

Viticulture

FEM experts will support companies in developing fully integrated, real-time solutions for smart monitoring of the winemaking process. Through its micro-vinification facility, consisting in temperature controlled tanks, sensors, and dataloggers, as well as enology laboratories, test and validation of artificial intelligence engines and visualization platforms will be supported, with the final goal to provide precise and accurate solutions for monitoring, with relevant impact on operational efficiency, as well as for driving the vinification precisely as desired, in terms of automation and repeatability of winemaking.