Initially developed at WebValley 2016 Summer School, Fruitipy aims at providing predictive analytics solutions for environmental health and food safety/quality from data gathered through portable spectrometers.
Fruitipy consists in a novel, open source web platform to support reproducible research with portable labware. Low-cost, portable spectrometers allow to monitor and collect data concerning fruit ripeness and quality in a non-invasive way. The data are sent to the database by means of an IoT infrastructure (smartphone/server), and then machine learning methods are applied to extract patterns and elaborate predictive models. Also Deep Learning algorithms are applied to the fruit photos in order to extract patterns and models based on fruit color.
Such a solution will allow to predict an improved harvest window for precision farming.
At the moment, Fruitipy is being applied to cultivations of grapes and berries, strawberries and late cherries, in collaboration with two major companies of Trentino, CAVIT wines and Sant’Orsola.