Wine4Cast - Spatio-temporal forecasting of wine productivity for multi-actor usability: integration of remote optical-photonic sensors and artificial intelligence
ONGOING
The aim of this project is to use an integrated approach involving the use of different parameters (weather, pollen, traditional counts) and a sensor technology component to create methods for predicting productivity, up to a year in advance, in some cases.
Wine4cast will use data from various sources such as atmospheric pollen, remote detection and phenology to train and evaluate machine learning models for detecting production components. In this project, the use of proximal optical-photonic sensors with different spectral resolutions, mounted on terrestrial or aerial platforms, will allow to support the winegrower with the digital phenotyping of production components. These proximal optical-photonic sensors provide the mapping of production components in order to forecast production at plot level, or even in intra-plot areas, from a precision viticulture perspective. It also provides production forecasting on a regional scale, by quantifying atmospheric pollen, followed by an innovative approach using remote sensing techniques (earth observation satellites, e.g. Sentinel) to determine the impact of post-flowering conditions on productivity.
Wine4Cast is a PRR project - R&D+i Projects - Agriculture 4.0", to be implemented between January 2023 and December 2025, coordinated by INESC TEC, in partnership with ADVID/CoLAB VINES&WINES, Duorum Vinhos, S.A., FCUP, Sociedade Vitivinícola Terras de Valdigem, S.A., Real Companhia Velha, S.A., Pólo de Inovação de Nelas, SPIN.WORKS, COTHN and Francisco Donas-Botto Rodrigues.
The aim of this project is to use an integrated approach involving the use of different parameters (weather, pollen, traditional counts) and a sensor technology component to create methods for predicting productivity, up to a year in advance, in some cases.
Wine4cast will use data from various sources such as atmospheric pollen, remote detection and phenology to train and evaluate machine learning models for detecting production components. In this project, the use of proximal optical-photonic sensors with different spectral resolutions, mounted on terrestrial or aerial platforms, will allow to support the winegrower with the digital phenotyping of production components. These proximal optical-photonic sensors provide the mapping of production components in order to forecast production at plot level, or even in intra-plot areas, from a precision viticulture perspective. It also provides production forecasting on a regional scale, by quantifying atmospheric pollen, followed by an innovative approach using remote sensing techniques (earth observation satellites, e.g. Sentinel) to determine the impact of post-flowering conditions on productivity.
Wine4Cast is a PRR project - R&D+i Projects - Agriculture 4.0", to be implemented between January 2023 and December 2025, coordinated by INESC TEC, in partnership with ADVID/CoLAB VINES&WINES, Duorum Vinhos, S.A., FCUP, Sociedade Vitivinícola Terras de Valdigem, S.A., Real Companhia Velha, S.A., Pólo de Inovação de Nelas, SPIN.WORKS, COTHN and Francisco Donas-Botto Rodrigues.
