• High Technology Service
    for Preventing Attacks
    of Fungal Diseases in Viticulture

Technical Development

Steps to take

picture

Having concluded its first prototype phase, the DecaVelta project is articulated according to the following by now consolidated elements:

1. ADCON Telemetry UNITS of radio or GSM type. Such a unit is made-up of a central unit associated with one or more secondary units in relation to the type of vineyard or of orography of the soil.
2. The unit allows the monitoring of the following environmental parameters with minimum 15 minutes interval periods: Quantity of precipitation, Soil temperauture, Sun irradiation, Relative humidity of the atmosphere, Wind speed, Direction of the wind, Temperature of the astmosphere, Leaf wetness, Soil humidity at 20cm, Soil humidity at 40cm, Further data services such as: battery charge level, solar panel power supply, and status of the connection.
3. Possibility of viewing the parameters via Internet, both through Workstation as well as on Pocket PC connected via GPRS
4. Specific Software of data processing information with the main algorithms implemented for the prediction of downy mildew infestation (Goidanich)
5. Possibility of exporting sets of data from internet client for further purposes, both in Microsoft Excel and XML format
6. As soon as further historical info will be available in the course of the project, revisionary functionalities of diseases based on non-linear models (Neural Network)

Will be implemented The system learns automonously as it accumulates intermediate data results. Through a rigorous workflow which includes: Data Sending, Data Warehouse, Data Mart, Data Mining & Artificial Intelligence, and Procedures & Field Operations, prediction charts are obtained which assert the services offered.

In the case of the software offered for the prevention of downy mildew infestation, it won’t be complex to interact with the Neural Network since this will provide data exclusively concerning possible risks of infestation. The final user will never have to deal with the complexity of the neural networks. He/she will exclusively benefit from their predictive abilities. The data can be viewed and/or entered also via Pocket PC.