What presented so far is the result of a preliminary study, based on only two vintages and resulting from their specific climatic trend. In particular, the high rainfall of spring 2019 surely weighed on the number of treatments needed against downy mildew, regardless of the use of the novel technology. The experimental design of the project, dedicated to a first collection of the field data necessary for the calibration of the forecast model, led to the hypothesis on the reductions in the cost of defense and environmental impact on the basis of ex-post evaluations. A more exhaustive study should be produced with field validation tests in which the infection risk forecasts provided by the model are directly adopted as a defense strategy by the companies. This may highlight not only a reduction in treatments but, above all, how the correct positioning of the same can make the difference in the agronomic management. It’s interesting, for example, noting how some cases of epidemic spread of downy mildew infections in 2019, on treated vineyards, were due to treatments carried out by the company too early with respect to the rains and to the very fast growth of vegetation, in the early stages of the season. The progressive refinement of the forecast model, using the data already collected and the comparison with new seasons, will significantly improve the presented results.

Plasmopora viticola is the agent of downy mildew of the grape vine, one of the most dangerous diseases in worldwide viticulture. The prevention of such a disease requires the frequent use of pesticides, which are usually sprayed according to a personal perception of the infection risk, not supported from objective data from the field. The GO PEI PVsensing propose a new system to improve the phytosanitary management of the vineyard against P. viticola, based on the adoption of novel sensors which are collecting climatic and environmental parameters from the fields, some of which have never been directly measured in the field before. Those parameters are the input of a previsional model, which computes the risk of an infection outburst in the crop, potentially with much higher accuracy than the existing systems. The final objective is to guide the farmer with a precise and reliable tool for the optimization of pesticide spraying, performing it more rationally, thus in the number and with the dosages strictly necessary for the protection of the crop, avoiding wastes that are not  motivated by a real risk of infections and, on the other side, are polluting the environment.

A total of 11 farms (both conventional and organic) will participate in the field experimentation of the system.

The accuracy enhancement of the system due to the new sensors will be assessed by a detailed agronomic analysis and there will be an evaluation of the environmental impact and the cost-benefit ratio.