In agriculture, increasingly strict regulatory restrictions on the pesticide residues in crops and the desire to protect the environment present a challenge of optimizing spraying to become as efficient as possible while minimizing the use of protective measures. The thesis deals with designing a prototype of an integrated system for automatic alerting for the need for spraying. In order to predict favorable dates for spraying, we model and predict the development of the observed pest in a particular area. Unlike typical pest development modeling approaches, our solution automatically obtains relevant data from an existing pest monitoring system in real time, calculates a pest development forecast and suggests the most favorable dates for spraying. The forecast is performed within a designed prototype of a decision support system that provides a user-friendly overview and forecast of potential favorable dates for spraying. The advantage of our solution is also the integration with an automatic alerting system, which offers immediate alerting based on spraying forecasts and the review of alerts on the field using a designed mobile application.
|
